Software. FP7 Project Portfolio. & Services. Cloud Computing, Internet of Services & Advanced Software Engineering

Digital Agenda for Europe

Software

&

Services

FP7 Project

Portfolio

Objective ICT-2011.1.2

Call 8 of FP7-ICT

June 2012

Cloud Computing,

Internet of Services

& Advanced Software

Engineering

LEGAL NOTICE

By the Commission of the European Union, Communications Networks, Content & Technology Directorate-General, Software & Services, Cloud Computing Unit

Neither the European Commission nor any person acting on its behalf is responsible for the use which might be made of the information contained in the present publication.

The European Commission is not responsible for the external web sites referred to in the present publication.

The views expressed in this publication are those of the authors and do not necessarily reflect the official European Commission view on the subject.

ISBN 978-92-79-25182-5 doi:10.2759/85125 © European Union, 2012

Reproduction is authorised provided the source is acknowledged.

Table of contents

Objective 1.2 Cloud Computing, Internet of Services and

Advanced Software Engineering……….. 4

Introduction………... 6

ARTIST………...10

BET aas………... 12

Bigfoot……… 14

Broker@Cloud... 16

CELAR………18

CloudScale……….. 20

Cloudspaces……… 22

COMPOSE………. 24

HARNESS……….. 26

LEADS……… 28

MARKOS………30

MIDAS……… 32

MODACLOUDS……… 34

OCEAN……….. 36

OpenI……….. 38

OSSMETER………... 40

PaaSage………... 42

PROSE……… 44

PROWESS………. 46

RISCOSS……… 48

SUCRE……… 50

U-QASAR……….. 52

ICT Work programme 2011 - 2012

Objective ICT-2011.1.2

Cloud Computing, Internet of Services and

Advanced Software Engineering

Target outcomes

a) Cloud Computing

- Intelligent and autonomic management of cloud resources, ensuring agile elastic scalability. Scalable data management strategies, addressing the issues of heterogeneity, consistency, availability, privacy and supporting security.

- Technologies for infrastructure virtualisation, cross platforms execution as needed for service composition across multiple, heterogeneous environments, autonomous management of hardware and software resources.

- Interoperability amongst different clouds, portability, protection of data in cloud environments, control of data distribution and latency.

- Seamless support of mobile, context-aware applications.

- Energy efficiency and sustainability for software and services on the cloud.

- Architectures and technologies supporting integration of computing and networking environments; implications of Cloud Computing paradigm on networks.

- Open Source implementations of a software stack for Clouds. b) Internet of Services

- Service engineering principles, methods and tools supporting development for the Internet of Services, including languages and tools to model parallelism.

- Services enabled by technologies for seamless integration of real and virtual worlds, through the convergence with Internet of Things and Internet of Contents.

- Massive scalability, self-management, verification, validation and fault localisation for software-based services.

- Methods and tools to manage life cycle of secure and resilient Internet-scale applications from requirements to run-time and their adaptive evolution over time.

c) Advanced software engineering

- Advanced engineering for software, architectures and front ends spanning across all abstraction levels.

- Quality measure and assurance techniques which adapt to changing requirements and contexts, to flexibly deal with the complexity and openness of the Future Internet. - Management of non-functional requirements typical of Internet-scale applications, like

concurrency levels which will be orders of magnitude larger than in today's applications, huge data stores and guaranteed performance over time.

- Tools and methods for community-based and open source software development, composition and life cycle management.

d) Coordination and support actions

- Support for standardization and collaboration in software and services technologies. - Support for the uptake of open source development models in Europe and beyond.

- Collaboration with Japanese entities on: cloud computing, particularly on common standards for data portability and on interoperability; services having more efficient energy usage.

Expected impact

• Emergence of European interoperable clouds contributing to an internal market of services in the EU whilst providing very significant business opportunities to SME's; improved trust in cloud-based applications and storage for citizens and business.

• Availability of platforms for easy and controlled development and deployment of value-added services through innovative service front-ends.

• Lower barriers for service providers and users to develop, select, combine and use value-added services through significant advances in cloud computing technologies and standardised and open interfaces.

• Efficient implementation of mainstream software applications on massively parallel architectures.

• Easier evolution of legacy software over time, thanks to innovative methods and tools managing the complete lifecycle of software from requirements to run-time.

• Fast innovation cycles in service industry, e.g. through the use of Open Source development model.

• A strengthened industry in Europe for software-based services offering a large choice of services satisfying key societal and economical needs, with reinforced capabilities to engineer and produce software solutions and on-line services.

Introduction

Objective ICT-2011.1.2 "Cloud Computing, Internet of Services

and Advanced Software Engineering"

The objective focuses on technologies specific to the networked, distributed dimension of software and access to services and data. It supports long-term research on new principles, methods, tools and techniques enabling software developers in the EU to easily create interoperable services based on open standards, with sufficient flexibility and at a reasonable cost.

The call was structured around three research oriented target outcomes: a) Cloud Computing, b) Internet of Services, and c) Advanced Software Engineering. An additional target outcome focussed on coordination and support activities.

Cloud Computing

Within the target outcome “Cloud Computing”, a number of topics were called for. Intelligent and autonomic management of cloud resources, ensuring agile elastic scalability is addressed by PaaSage. Its open and integrated platform will allow users to manage Cloud resources and to have autonomic support at execution time to optimise application deployment. CELAR will dynamically allocate and de-allocate resources required by applications to be executed over cloud platforms, thanks to an intelligent decision-making module, which decides on the optimal expansion or contraction of allocated resources and their type in real-time. LEADS will provide the automatic management of CPU, network, and data resources across multiple clouds and it will provide elastic scalability across multiple levels of abstractions, i.e. providing agile vertical scaling mechanisms at the VM level and agile horizontal scaling within a cloud and across multiple clouds.

Scalable data management strategies, addressing the issues of heterogeneity, consistency, availability, privacy and supporting security is addressed by BigFoot that will develop a novel, scalable system for processing and interacting with large volumes of data, tackling issues related to scalability, data and storage heterogeneity. CloudSpaces deals with consistency and replication over heterogeneous repositories in Personal Clouds.

Technologies for infrastructure virtualisation, cross platforms execution as needed for service composition across multiple, heterogeneous environments, autonomous management of hardware and software resources is addressed by HARNESS and – to a lesser extent – CELAR. HARNESS will study new virtualisation techniques tailored to heterogeneous hardware and network resources for flexible allocation and reallocation and advanced techniques for optimising cost/performance allocation trade-offs. CELAR offers an innovative approach for autonomous management of hardware and software resources.

Interoperability amongst different clouds, portability, protection of data in cloud environments, control of data distribution and latency is addressed by PaaSage, LEADS, and – to a lesser extent – CloudSpaces. PaaSage will allow dynamic interoperability among different Clouds, and transparent simultaneous use of both private Clouds and commercial Clouds, thanks to its Cloud modelling language and PaaSage platform. LEADS will ensure portability across different clouds by wrapper libraries, and it will provide data protection via encryption and access control mechanisms. CloudSpaces will study semantic interoperability of Personal Cloud data, as well as privacy-aware data sharing among different Personal Clouds.

OpenI will develop technical infrastructures that enable the sharing of data and content across a user’s applications, devices and platforms. LEADS will use a novel approach of energy-aware scheduling in the cloud.

The last Cloud Computing topic is Open Source implementations of a software stack for Clouds, which is a secondary objective of several projects. BigFoot aims to deploy a novel, scalable system for processing and interacting with large volumes of data as a full-fledged software stack for private cloud environments. OpenI will deliver a Cloud platform under an open source governance model, entirely built on open cloud stacks, and compatible from the start with existing cloud hosting providers that support open cloud technologies. Building on past/current projects that have contributed to a software stack for a Cloud infrastructure, PaaSage will offer a multi-cloud development platform that will work on top of existing infrastructure software, complementing what is currently available and completing the open source software stack for Clouds.

Internet of Services

Within the target outcome “Internet of Services”, service engineering principles, methods and tools supporting development for the Internet of Services, including languages and tools to model parallelism are addressed by Broker@Cloud and – to a lesser extent – by CloudScale, COMPOSE, and ARTIST. Broker@Cloud considers the model of Cloud Service Brokerage emerging as a central component of the future Internet of Services, and therefore places the implications of this model with respect to principles, methods and tools for service management as its central topic of investigation. CloudScale provides engineering principles, methods and tools that support the development of scalable services and service compositions. COMPOSE provides methods and tools necessary to discover, integrate, combine, execute, and manage highly distributed services, objects, and content. ARTIST will provide a complete solution with methods, principles and tools supporting the adaptation of legacy applications to the Future Internet, and preparing them for being offered as SaaS over the Internet.

Services enabled by technologies for seamless integration of real and virtual worlds, through the convergence with Internet of Things and Internet of Contents are well addressed by COMPOSE and BETaaS. COMPOSE will provide support for integrating the Internet of Things, the Internet of Contents, and the Internet of Services, ensuring smooth convergence of the real towards virtual worlds. To this end, COMPOSE will support seamless virtualisation of existing content provisioning systems and objects into semantic services that can better be discovered, integrated, combined, executed, and managed. BETaaS will lay the foundations of fast and cost-effective development of Machine-to-Machine applications, also providing an environment for their efficient execution. The BETaaS platform will provide a set of services smart things in the Internet of Things with their representation in the Internet of Contents.

Massive scalability, self-management, verification, validation and fault localisation for software-based services are addressed by CloudScale and – to a less extent – by BETaaS. CloudScale provides methods and tools to better achieve massive scalability of new and legacy software services. Because of the envisaged growth of the Internet of Things, massive scalability is a key requirement for future adoption of BETaaS, which will use a fully distributed architecture hosted on the Machine-to-Machine gateways. Furthermore, BETaaS addresses self-management through the use of semantic technologies to model the behaviour of things.

Methods and tools to manage life cycle of secure and resilient Internet-scale applications from requirements to run-time and their adaptive evolution over time are addressed by COMPOSE and Broker@Cloud. COMPOSE will provide a marketplace and supporting infrastructure for the discovery, creation, combination, and exploitation of highly adaptive and distributed added-value Internet services built on top of existing content, objects, and services. Broker@Cloud will deliver methods and mechanisms helping brokers to address the issue of cloud service lifecycle management including keeping track of the evolution of services over time and understanding the impact that changes can have on different consumers.

Advanced Software Engineering

Within the target outcome “Advanced Software Engineering” a number of topics were called for.

Advanced engineering for software, architectures and front ends spanning across all abstraction levels is well addressed by MODAClouds and ARTIST. MODAClouds will deliver an advanced model-based approach and an Integrated Development Environment to support system developers in building and deploying applications, with related data, to multi-Clouds spanning across the full Cloud stack (IaaS/PaaS/SaaS). ARTIST will provide tools for a Reverse Engineering process that will extract all necessary information from an original legacy application and will produce models of the latter. Furthermore, through Forward Engineering techniques and tools, ARTIST will produce final models of a service-oriented application, at different levels of abstraction, complexity and concerns, in order for this to be deployed in a target cloud environment.

Quality measure and assurance techniques which adapt to changing requirements and contexts, to flexibly deal with the complexity and openness of the Future Internet is very well addressed by several prioritised proposals. PROWESS aims to automate quality assurance, reducing its cost and improving effectiveness, based on properties of the system that should hold. It will provide a development process and tools that ensure dependable quality of service through directly verifying high-level properties of a system. U-QASAR will create a flexible Quality Assurance, Control and Measurement Methodology to quantify the quality of Internet-related software development projects and their resulting products. In addition to design-time and run-time quality measures and assurance techniques, MODAClouds will offer prediction mechanisms and adaptive policies for addressing automatically specific critical cases. MIDAS will provide an integrated framework for SOA testing automation that spans all testing activities: test generation, execution, evaluation, planning and scheduling, on the functional, interaction, fault tolerance, security and usage-based testing aspects. As for Management of non-functional requirements typical of Internet-scale applications, like concurrency levels which will be orders of magnitude larger than in today’s applications, huge data stores and guaranteed performance over time, MIDAS will offer security testing as well as fault tolerance testing, i.e. testing of a system’s capability to cope with the unavailability of resources necessary to deliver/use services.

Tools and methods for community-based and open source software development, composition and life cycle management are well addressed by several prioritised proposals. MARKOS will identify the relationship between components located at different forges or sites, and simplify the discovery of Open Source components distributed on the Net through software metadata published as Linked Data with a uniform and central access. RISCOSS will offer novel risk identification, management and mitigation tools and methods for community-based and industry-supported Open Source Software (OSS) development, composition and lifecycle management to individually, collectively and/or collaboratively manage OSS adoption risks. OSSMETER targets automated analysis and measurement of open-source software, and will develop a platform that will support decision makers in the process of discovering, comparing, assessing and monitoring the health, quality, impact and activity of OSS. It will compute trustworthy quality indicators by advanced analysis and integration of

information from diverse sources including project metadata, source code repositories, communication channels and bug tracking systems of OSS projects.

Coordination and Support Actions

As for Coordination and Support Actions, three topics were called for. Support for standardization and collaboration in software and services technologies is addressed by OCEAN, which will analyse the current status and standardisation needs for the emergence of an Open Cloud in Europe. OCEAN’s goal is to foster the emergence of a sustainable open source cloud offering and boost market innovation in Europe, by generating greater efficiency and economies of scale among European FP7 collaborative research projects on open source cloud computing.

Support for the uptake of open source development models in Europe and beyond is addressed by PROSE, OCEAN, and SUCRE. PROSE is principally concerned with providing a platform to coordinate FLOSS from FP7 projects. OCEAN will help FP7 research projects on Cloud Computing to spot open source projects with which they could collaborate and to identify what software components they could use to achieve their objectives. SUCRE will explore the adoption of Cloud Computing coupled with the use of Open Source development models in the Public Sector and the Health Care Industry. It will also operate an expert group with industrial representatives from Europe and Japan to discuss, amongst others, the uptake of open-source development models.

Objective ICT-2011.1.2 "Cloud Computing, Internet of Services and Advanced Software Engineering"

Note: some projects may appear in several areas

More detailed factsheets of each project are provided in the remainder of this brochure. The factsheets are listed in alphabetical order of the projects' acronyms.

ARTIST

Concept

The Cloud-based service delivery model has the potential to create tremendous new business opportunities for software companies. On-going improvements in the Internet’s connections, both in speed and reliability as well as reach, have made Internet native solutions an attractive alternative, and the rate of innovation driving software and service evolution is still accelerating. Innovations in the technological space affect the systems that the software has to support or needs to adapt to. Innovations in the business space not only affect the licensing and usage model, but also the core value proposition to the customer. To remain viable, legacy software solutions have to be improved with regard to these new circumstances, but without disrupting the business continuity of existing customers. Software service companies need to transition to a new opportunity model, without abandoning their client portfolios. The complete lifecycle of software, from requirements to delivery and operations, has to be re-adapted to the new technological and business conditions, requirements and challenges. There is a need for tools and methods to support software evolution and adaptation as a key value for next generation service-based software modernization. Following this approach, companies face the following challenges:

ƒ The decision whether to migrate their existing products or to start from scratch;

ARTIST proposes a software modernization approach based on Model Driven Engineering techniques to automate the reverse and forward engineering of legacy applications to and from platform independent models. It reduces the risk, time and cost of migrating legacy software and lowers the barriers for service companies wanting to take advantage of the latest Cloud Computing and Software as a Service based technologies and business models.

AT A GLANCE

Project title:

Advanced software-based seRvice provisioning and migraTIon of legacy SofTware

Project reference:

317859

Project coordinator:

Clara Pezuela, ATOS Spain SA, SPAIN

Partners:

Fundacion Tecnalia Research & Innovation, (ES) Inria - French National Institute For Research In Computer Science and Control, (FR)

Fraunhofer-Gesellschaft zur Foerderung der Angewandten Forschung E.V., (DE)

Technische Universität Wien, (AT) Engineering, (IT)

Institute Of Communication and Computer Systems, (GR)

Sparx Systems Software GMBH, (AT) Athens Technology Center SA, (GR) Spikes NV, (BE)

Duration:

Total cost:

Website:

ƒ The estimation of the impact and effort required to implement the modernization of a system is difficult and uncertain; ƒ Time-to-market is critical. Therefore the

software development cycles need to change;

ƒ High requirements for specialized skills due to a low degree of process automation.

A complete approach is needed that helps companies bring their applications and services into the Internet of Services, taking into account the implications of current architectures, and forecasting the implications of future ones. This requires the development of a new vendor and platform independent methodology and a new automation oriented toolset for reengineering, migration, maintenance and evolution.

This is the mission of ARTIST.

Objective

To prepare, support and increase the competitiveness of the European Software and Services Industry in a global Cloud and Software as a Service (SaaS) business environment, ARTIST develops a set of methods, tools and techniques that facilitate the transformation and modernization of legacy software assets and businesses. The project creates tools to assess, plan, design, implement and validate the automated evolution of legacy software to SaaS and the Cloud Computing delivery model. By focusing on reusability during this transition, the methods and tools are generic enough to cover future shifting efforts, e.g. deployment to future platform delivery paradigms.

Approach

In order to reach its objective, ARTIST: ƒ Develops an innovative and combined

technical and business analysis on the maturity and prospect of the legacy application;

ƒ Provides a large scale model-based approach for representingthe source and target applications as well as infrastructures/platforms;

ƒ Creates a unified performance modelling framework;

ƒ Identifies dynamic deployment methodologies;

ƒ Fosters reusability of the (modelling) artefacts produced during the migration process through the usage of a repository; ƒ Implements an innovative and thorough

testing and continuous validation process that will span across all layers of the multi-level ecosystem;

ƒ Enhances, enforces and promotes the usage of an integrated certification model for Cloud application providers.

Impact

Easier evolution of legacy software over time, thanks to innovative methods and tools managing the complete lifecycle of software from requirements to run-time.

This major impact is expected to be produced by ARTIST through allowing software vendors and users of open source software to migrate their legacy software to a new software paradigm in an automated, easy and cost effective way. This means that legacy software can be transformed so that it receives the benefits of that new software paradigm such as performance enhancement, cost effectiveness, and better interoperability.

BETAAS

Motivation

Today there are countless devices at work to improve productivity and quality of life of human beings, in all technological domains. In most cases they operate in isolation or with very little cooperation from their likes, and serve a well-defined single purpose for which they have been engineered. Such situation is sub-optimal because: (i) fine-grained (raw) data have to be conveyed in a centralized manner over the Internet from sensors up to the remote center, thus, the things and gateways are effectively separated from the back-end both physically and logically; (ii) the current approach is vertical, i.e. each M2M application has its own remote center for data storage and processing.

Concept

To overcome the limitations of the current systems for M2M applications, we propose a novel approach based on the following principles:

1. Storage and processing of data are as close as possible, in space and time, to where they are generated and consumed.

2. Important non-functional requirements, namely security, reliability and QoS, are supported by means of a tight integration of the high-level services provided with the physical resources of the peripheral devices, i.e. things and gateways.

3. Energy efficiency and scalability of the systems are achieved through the distribution

In BETaaS a platform will be developed for the execution of machine-to-machine (M2M) applications built on top of services deployed in a “local cloud” of gateways. Scalability, security, dependability, context and resource awareness, and quality of service (QoS) will be embedded “by design” into the platform. Validation will be done through experiments targeting the Smart City and Home Automation use cases.

AT A GLANCE

Project title:

Building the Environment for the Things as a Service

Project reference:

Project coordinator:

Partners:

ATOS, (ES)

Hewlett-Packard Italy, (IT)

Converge ICT Solutions and Services, (GR) Tecnalia, (ES)

Aalborg University – CTIF, (DK) Università di Pisa, (IT)

Duration:

Total cost:

Website:

of on-the-spot inferred content, rather than raw data.

These principles will be realized by defining a content-centric platform distributed over a local cloud, hosted by the gateways, providing an environment for applications accessing M2M services and devices through a set of services. Its deployment will be dynamic so as to follow the time-varying M2M services and the changing characteristics of the applications.

Project goal

To design and realize a runtime platform for the deployment and execution of content-centric M2M applications, which relies on a local cloud of gateways.

The proposed platform will provide a uniform interface and services to map content (information) with things (resources) in a context-aware fashion.

Deployment of services for the execution of applications will be dynamic and will take into account the computational resources of the low-end physical devices used.

To this aim, the platform will need to be based on a suitable defined Internet of Things model, which will allow the integration of the BETaaS components within the future Internet environment.

Impact

The barriers to enter the M2M segment today are high, since the market is very fragmented and user requirements and expectations are very heterogeneous. This scenario is especially harsh for small and medium players, who do not have a sufficiently vast commercial influence or the strength for massive marketing campaigns. By defining open interfaces, which will be also pushed for standardization in the relevant technical committees, M2M service providers have reduced risks of the investment and less training/setup costs, since they can re-use the skills and experience over different projects and also across various domains.

The BETaaS platform will be released as open source to achieve the following goals: i) to benefit from the contributions of the open source community of developers (in terms of customization, testing, developments… ); ii) to allow M2M service providers to focus on the application-specific aspects, without the need for working on common features, which reduces the development costs and time-to-market.

BIGFOOT

The Context

The amount of digital information in our world has been exploding and new technologies and services will continue to fuel exponential growth of large pools of data that can be captured, stored, and analyzed. Nowadays, however, tools and services to store, process and interact with data are still in their infancy, represented by scattered solutions that fall short in having a unified vision, that lack common interfaces, and that only offer best-effort services. The aim of BigFoot is to overcome current drawbacks by designing, implementing and evaluating a Platform-as-a-Service. The BigFoot stack features automatic and self-tuned deployments of storage and processing services for private clouds, going beyond best-effort services currently available in the state-of-the-art. BigFoot takes a novel, cross-layer approach to system optimization, which is evaluated with a thorough experimental methodology using realistic workloads and datasets from two representative applications, namely ICT Security and Smart Grid data analytics. In addition, BigFoot aims at making data interaction easy by supporting high-level languages and by taking a service-oriented approach to support and optimize latency sensitive queries.

The aim of BigFoot is to design, implement and deploy a Platform-as-a-Service solution for processing and interacting with large volumes of data coming from ICT Security, Smart Grid and other application areas. The BigFoot stack builds upon and contributes to the Apache Hadoop ecosystem and the Apache OpenStack project.

AT A GLANCE

Project title:

Big Data Analytics of Digital Footprints

Project reference:

Project coordinator:

Partners:

Ecole Polytechnique Federale de Lausanne, (CH) Technische Universität Berlin, (DE)

Symantec Lab, (IE) CEO, GridPocket SAS, (FR)

Duration:

Total cost:

Website:

WP3 Batch Analytics and Interactive Query Engines

WP2

WP4

WP5 Applications

Virtualization Layer Distributed Data Stores

Hadoop Distributed Filesystem Infrastructure Virtualization High-level query to low-level program translation Optimization Library Service-oriented Query Engine Distributed Datastore Data Partitioning and

Placement

Data Store Support

Service Deployment Service Optimization Data mining Algorithms Statistical Patterns Analysis Interactive Queries Virtualization Layer 1 2 3 Hadoop MapReduce Compiler BigFoot So ft ware Stack

BigFoot Approach

BigFoot merges several research domains, including the design of scalable algorithms for data mining, large-scale, fault-tolerant distributed systems, and virtualization technologies. These areas blend together resulting in a full-fledged software stack, illustrated by the Figure below, and described as follows:

The Virtualization Layer includes

mechanisms for the virtualization of the BigFoot infrastructure (including server machines and network), the deployment of BigFoot instances, and their optimization. The Data Stores Layer takes care of the optimization of data partitioning and placement – in conjunction with the virtualization layer – and offers high-availability support to overcome SPOFs.

The Batch and Interactive Engines

Layer implement several data-flow

optimizations and provide a new approach to express batch data analysis using a high- level language. The service-oriented query engine is a novel component to optimize and facilitate the interaction with data.

The BigFoot Impact

Several sources suggest that we are on the cusp of a tremendous wave of innovation, productivity, and growth as consumers, companies, and economic sectors exploit the potential of large amounts of data directly or indirectly produced by their interaction. The aim of BigFoot is to produce:

High quality research, by focusing outstanding problems in the domain of large-scale data management, with an approach that blends systems and theory research. The cross-layer approach to system optimization brings additional challenges that have not yet been addressed in the literature. Finally, measurement-based analyses of typical workloads (as obtained from real-world applications and data) represent an invaluable contribution to the research community.

Open-source Software, by contributing the BigFoot software stack and a selection of its inner components to relevant

communities. In particular, BigFoot builds upon the Apache Hadoop ecosystem and the Apache OpenStack project: as such, BigFoot uses the Apache Software Foundation v2.0 License. BigFoot uses a two-pronged approach to software development: it maintains public GIT repositories for experimental-level code – which are incrementally populated – and it uses the appropriate ticketing (JIRA) system to contribute to existing projects.

Experimental Platforms, to run thorough, reproducible comparative analyses of system performance as achieved by the BigFoot stack and other existing (even commercial) solutions to Big Data analysis. BigFoot will contribute to Symantec’s WINE initiative, where it is going to be deployed as a PaaS, and to GridPocket’s energy services platform.

Broker@Cloud

Motivation

As enterprises increasingly adopt the model of cloud computing, their IT environments are transformed into a matrix of interwoven infrastructure, platform and application services delivered by multiple providers. In order to deal with the complexity of consuming large numbers of cloud services from diverse sources, enterprises will need assistance from specialised cloud service delivery intermediaries. These will need to offer an array of sophisticated brokerage services which will go far beyond the kinds of intermediation capabilities available today.

The challenge

The challenge taken up by Broker@Cloud is to research and to develop solutions with respect to some of the most valuable and technically demanding types of brokerage capabilities foreseen for future enterprise cloud service brokers.

We envisage developing a brokerage framework which will allow cloud intermediaries to equip their platforms with advanced methods and mechanisms for continuous quality assurance and optimization of software-based cloud services. Those software-based services can range from simple programmatically-accessible web APIs, to complex software applications delivered as cloud services, i.e. on-demand Software-as-a-Service offerings.

The goal of the Broker@Cloud project is to develop a framework that will equip cloud service intermediaries with advanced methods and mechanisms for continuous quality assurance and optimization of software-based cloud services. The framework will allow enterprise cloud service brokers to monitor the obligations of providers towards consumers, as well as to detect opportunities for optimising service consumption.

AT A GLANCE

Project title:

Enabling Continuous Quality Assurance and Optimization in Future Enterprise Cloud Service Brokers

Project reference:

Project coordinator:

Partners:

CAS Software, (DE)

Institute of Communication and Computer Systems, (GR)

SAP, (DE)

South-East European Research Centre, (GR) SingularLogic, (GR)

The University of Sheffield, (UK)

Duration:

Total cost:

Website:

Employing the capabilities provided by the Broker@Cloud framework will assist future enterprise cloud service brokers in providing assurances towards consumers with respect to how reliable and how optimal the delivered services are.

The goal is to provide future cloud intermediaries with advanced means of monitoring both the obligations of each cloud service provider towards consumers (as well as towards the intermediary itself), and the opportunities for optimising the services each consumer receives, as soon as these surface.

Broker@Cloud framework

The brokerage framework, most of which will be released as Open Source Software, will comprise the following core building blocks:

1. Capabilities for cloud service governance and quality control (lifecycle management, dependency tracking, policy compliance, SLA monitoring, certification testing)

2. Capabilities for cloud service failure prevention and recovery (event monitoring, reactive and proactive failure detection, adaptation analysis and

recommendation)

3. Capabilities for continuous optimization of cloud services (optimisation opportunity detection and analysis based on cost, quality, or functionality preferences)

4. Interfaces and methods for platform-neutral description of enterprise cloud services (technical, operational and business aspects, static and dynamic views)

The validation of project results will be done through two pilot case studies, during which we will integrate the Broker@Cloud framework into two different enterprise cloud service delivery platforms (CAS Open and SingularLogic Galaxy).

Impact

The results of Broker@Cloud are expected to be of significant value to the enterprise software industry, which is presently hard-pressed to understand the implications of the ongoing paradigm shift towards cloud computing, and to redefine roles and opportunities in the emerging setting. Dissemination and an open source development model is the strategy through which the consortium will seek to achieve this impact.

CELAR

Project Rationale - the idea

behind CELAR

Auto-scaling resources is one of the top obstacles and opportunities for Cloud Computing: consumers can minimize the execution time of their tasks without exceeding a given budget; cloud providers maximize their financial gain while keeping their customers satisfied and minimizing administrative costs. Many systems claim to offer adaptive elasticity, yet the “throttling” is usually performed manually, requiring the user to figure out the proper scaling conditions. In order to harvest the benefits of elastic provisioning, it is imperative that it is performed in an automated, fully customizable manner. CELAR delivers a fully automated and highly

customizable system for elastic

provisioning of resources in cloud computing platforms.

CELAR Technical Realization

The main outcome of the project is a complete set of open-source tools that will allow the enhancement of a Cloud platform towards automatic, intelligent, multi-grained resource provisioning according to the needs of user applications. Specifically, the CELAR contribution comprises four main parts: (i) the elasticity provisioning subsystem, which is a middleware that adaptively and automatically manages platform resources;

(ii) the c-Eclipse framework to provide plug-ins for accessing and managing Cloud resources on the envisioned platform; (iii) a Cloud information subsystem, which contains a Cloud resource description framework and search capabilities for Cloud

The CELAR project provides automatic, fine-grained resource allocation for Cloud applications. This enables the commitment of just the right amount of resources based on application demand, performance and requirements, resulting in optimal use of infrastructure and significant reductions in costs.

AT A GLANCE

Project title:

CELAR: Cloud ELAsticity pRovisining

Project reference:

317790

Project coordinator:

Innovation Centre in Information, Communication and Knowledge Technologies, GREECE

Partners:

University of Cyprus, (CY)

Vienna University of Technology, (AT) GRNET S.A., (GR)

Playgen, (UK)

Institute of Cancer Research, (UK) Sixsq Sarl, (CH)

Flexiant Limited, (UK)

Duration:

Total cost:

Website:

resources; (iv) a scalable, multi-layer Cloud Monitoring tool that gathers a rich set of platform, infrastructure and application-side metrics and evaluates them in a composite fashion. These modules are both generic in nature and open-source, in order to allow for maximum utilization and ease of adaptation with existing commercial, academic and community systems. Providing added value and simplifying application deployment over CELAR, the project also develops a framework for the cloudification of any elasticity-demanding cloud application with the CELAR system, offering this integration into a single installable software package. The main outcome of the project (featuring a well-defined 3-layer architecture as described above) is depicted in the figure.

CELAR Impact and means of

achievement

CELAR actively contributes to two major goals defined by the digital agenda: interoperability and open access for ICT products and services. Its technology brings forth advantages that are manifold. The optimal use of cloud resources results in significant cost-reductions and increased application performance, required by both

infrastructure providers and users alike. Equally important, CELAR provides open, standardized access over the complete stack of the cloud ecosystem. This results inavoidance of vendor lock-ins, increased applicability and ease of service development and deployment, giving the project potential to become an invaluable European technological and economic asset.

Achieving the expected impact relies on three pillars:

1) Visibility for the developed technology 2) Meeting the needs of potential users

and offering new types of services to the respective customers.

3) Easing Adoption of the Technologyand clearly showcasing its potential benefits Two exemplary applications, in the areas

of online gaming and scientific

computing, showcase and validate the

developed technology, providing a clear path towards the adoption of the CELAR rationale and increasing the visibility and impact of the project.

CLOUDSCALE

The Problem

Cloud providers theoretically offer their customers unlimited resources for their applications on an on-demand basis. However, scalability is not only determined by the available resources, but also by how the control and data flow of the application or service is designed and implemented. Implementations that do not consider their effects can either lead to low performance (under-provisioning, resulting in high response times or low throughput) or high costs (over-provisioning, caused by low utilisation of resources).

Objectives

CloudScale provides an engineering approach for building scalable cloud applications and services. Our objectives: 1. Make cloud systems scalable by design

so that they can exploit the elasticity of the cloud, as well as maintaining and also improving scalability during system evolution. At the same time, a minimum amount of computational resources shall be used.

2. Enable analysis of scalability of basic and composed services in the cloud. 3. Ensure industrial relevance and uptake

of the CloudScale results so that scalability becomes less of a problem for cloud systems.

The goal of CloudScale is to aid service providers in analysing, predicting and resolving scalability issues, i.e. support scalable service engineering. The project extends existing and develops new solutions that support the handling of scalability problems of software-based services.

AT A GLANCE

Project title:

Scalability Management for Cloud Computing

Project reference:

317704

Project coordinator:

Partners:

SAP, (DE)

Ericsson Nikola Tesla, (HR) XLAB, (SI)

University of Paderborn, (DE)

Duration:

Total cost:

Website:

Results

CloudScale enables the modelling of design alternatives and the analysis of their effect on scalability and cost. Best practices for scalability further guide the design process. CloudScale provides tools and methods that detect scalability problems by analysing code. Based on the detected problems, CloudScale offers guidance on the resolution of bad practice. As a basis for all of this is a language (ScaleDL) that service providers use to specify the scalability properties of basic and composed cloud services.

Value Chain

CloudScale provides tools and methods supporting inherently massively scalable services architectures, enabling the European industry, including SMEs, to gain an advantage when developing services for the cloud.

The results of CloudScale are aimed at different types of people, organisations and roles, offering benefits to each:

• End users: Improved scalability of system deployed in clouds means satisfied users even during peak load.

• For developers of software services,

improved scalability management becomes a selling point. CloudScale tools help developers make sensible decisions about which parts of the system most require “gold plating”.

• System architects, the composers of software services are able to understand and predict the scalability of services resulting from compositions.

• Service providers are able to make timely decisions about purchase or deployment of more hardware in order to prevent scalability bottlenecks before they show up. They are also able to plan reduction in non-essential features to retain core functionality during periods of extreme demand.

IaaS (Infrastructure as a Service) providers may lose some business due to more effective use of resources by their customers, thanks to improved scalability. However, they are able to serve more customers with the same hardware through better management of scalability of their own systems, and can thereby operate with a smaller safety margin and greater profit.

CLOUDSPACES

Towards personal cloud 2.0

In the next few years, users will require ubiquitous and massive network storage to handle their ever-growing digital lives. In this line, the Personal Cloud model defines a ubiquitous storage facility enabling the unified and location agnostic access to information flows from any device and application. Popular providers like Dropbox or Ubuntu One already provide unified synchronization and sharing services to millions of users.

But Personal Clouds are in their infancy, and two major problems must be solved: First, there is a big privacy problem that precludes the adoption of this model by many users, companies and public institutions. Most Personal Clouds follow a simple centralized synchronization model that stores all information in the Cloud as a remote file system. The entire data management process is in the hands of the Cloud providers, so the users really lose control of where their information is stored and who can access it.

Another important problem is the lack of interoperability between Personal Cloud services impeding information sharing, but also precluding information portability among them. This generates what is known as vendor lock-in: a best decision now may leave a customer trapped with an obsolete provider later, simply because the cost to switch from one provider to another is prohibitively expensive.

The CloudSpaces project advocates for a paradigm shift from application-centric to user-centric models where users will retake the control of their information. To this end, CloudSpaces will devise an open service platform providing privacy-aware data sharing as well as interoperability mechanisms among heterogeneous Personal Clouds.

AT A GLANCE

Project title:

CloudSpaces: Open Service Platform for the Next Generation of Personal clouds

Project reference:

Project coordinator:

Pedro García López, Universitat Rovira i Virgili, SPAIN

Partners:

École Polytechnique Fédérale de Lausanne, (CH), Institut Eurecom, (FR),

Canonical Limited, (UK), eyeOS, (ES), TISSAT, (ES)

Duration:

Total cost:

Website:

CloudSpaces Platform

CloudSpaces aims to create the next generation of open Personal Clouds using three main building blocks: CloudSpaces Share, CloudSpaces Storage and CloudSpaces Services.

CloudSpaces Share will deal with interoperability and privacy issues. The infrastructure must ensure privacy-aware data sharing and trustworthy assessment from other Personal Clouds. It will also overcome existing vendor lock-in risks thanks to open APIs, metadata standards, personal data ontologies, and portability guarantees.

CloudSpaces Storage takes care of scalable data management of heterogeneous storage resources. In particular, users retaking control of their information implies control over data management. This new scenario clearly requires novel adaptive replication and synchronization schemes dealing with aspects like load, failures, network heterogeneity and desired consistency levels.

Finally, CloudSpaces Services provides a high level service infrastructure for third-party applications that can benefit from the

Personal Cloud model. It will offer data management (3S: Store, Sync, Share), data-application interfaces, and a persistence service to heterogeneous applications with different degrees of consistency and synchronization.

Impact

We are now in a decisive turning point that will definitely influence how we interact with the information in the following years. If the major players dominate this market with vertical walled garden solutions, there will be little space left for European Cloud providers, software solution providers and SMEs.

On the contrary, CloudSpaces foresees short time-to-market and important impacts. In particular, the project will reach a global impact thanks to our contributions to three open source projects with huge communities: Ubuntu One Personal Cloud, eyeOs Personal Desktop, and OpenStack Cloud middleware. Cloud providers will benefit from OpenStack Swift novel Personal Cloud services facilitating the emergence of European interoperable clouds.

End-users and companies will increase their trust in cloud-based applications and storage. This can ease the massive adoption of online services such as Ubuntu One or others. Software solution providers and SMEs will be able to build innovative services on top of open Personal Clouds. eyeOS Personal Desktop is our key proof of concept demonstrating the capabilities of the platform. CloudSpaces aims to create the next

generation of Personal Clouds, namely Personal Cloud 2.0, offering advanced issues like interoperability, advanced privacy and access control, and scalable data management of heterogeneous storage resources.

Furthermore, it will offer an open service platform for third-party applications leveraging the capabilities of the Open Personal Cloud.

COMPOSE

COMPOSE overview

COMPOSE aims at enabling new services that can seamlessly integrate real and virtual worlds through the convergence of the Internet of Services (IoS) with the Internet of Things (IoT).

COMPOSE will achieve this through the provisioning of an open and scalable marketplace infrastructure, in which smart objects are associated to services that can be combined, managed, and integrated in a standardised way to easily and quickly build innovative applications.

The project will develop novel approaches for virtualising smart objects into services and for managing their interactions. This will include solutions for managing knowledge derivation, secure and privacy-preserving data aggregation and distribution, dynamic service composition, advertising, discovering, provisioning, and monitoring.

COMPOSE is expected to give birth to a new business ecosystem, building on the convergence of the IoS with the IoT and the Internet of Content (IoC). The COMPOSE marketplace will allow SMEs and innovators to introduce new IoT-enabled services and applications to the market in a short time and with limited upfront investment.

At the same time, major ICT players, particularly cloud service providers and telecommunications companies, will be able to reposition themselves within a new IoT-enabled value chain.

COMPOSE will create an ecosystem for unleashing the power lying within the vast amount of internet connected smart objects by enabling easy construction of services based on these objects. COMPOSE technology will enable standardized access to such objects, the creation of base services, combine them into composite services, and finally building applications.

AT A GLANCE

Project title:

Collaborative Open Market to Place Objects at your SErvice

Project reference:

Project coordinator:

Partners:

CREATE-NET, (IT)

Fraunhofer Institute FOKUS, (DE) The Open University, (UK)

Barcelona Supercomputing Center, (ES) INNOVA S.p.A, (IT)

University of Passau, (DE) U-Hopper, (IT)

GEIE ERCIM (W3C), (FR)

Fundació Privada Barcelona Digital Centre Tecnològic (Bdigital), (ES)

Retevision, (ES) EVRYTHNG, (UK)

Duration:

Total cost:

Website:

Technical Approach

The vision of the COMPOSE project is to advance the state of the art by integrating the IoT and the IoC with the IoS through an open marketplace, in which data from Internet-connected objects can be easily published, shared, and integrated into services and applications. The marketplace will provide all the necessary technological enablers, organized into a coherent and robust framework covering both delivery and management aspects of objects, services, and their integration.

• Object virtualization: enabling the creation of standardized service objects • Interaction virtualization: abstract

heterogeneity while offering several interaction paradigms

• Knowledge aggregation: creating information from data

• Discovery and advertisement: of semantically-enriched objects and services

• Data Management: handle massive amounts and diversity of

data/metadata

• Ad hoc creation, composition, and maintenance: of service objects and services

• Security, heterogeneity, scalability, and resiliency: incorporated throughout the layers

Expected Impact

COMPOSE strives for a strong impact on a developing market by lowering barriers to

develop, select, combine, and use IoT-based standardized value added services. This will be achieved by providing a complete ecosystem, and having it adopted by enterprises, SMEs, government-related bodies, individual developers and end-users. Opening the door to this realm for smaller entities will lead to higher innovation. COMPOSE expects to aid by fostering a developers' community and advocating an open source/interfaces.

Use-Case Driven

COMPOSE design, development, and validation will be based on innovative use cases highlighting different aspects of the platform. Among the use cases:

• Smart City (Barcelona): Ample

amount and diversity of sensors are deployed at a Barcelona district under the supervision of a COMPOSE partner. Along with Barcelona's OpenData, COMPOSE intends to showcase life in a smart city by creating a group of city services for the citizens.

• Smart Territory (Trentino): With the collaboration of regional network providers, the tourism board, and meteorological data providers, COMPOSE will explore innovative services for tourists. This pilot aims to enhance the tourist experience by exploiting COMPOSE technologies for the creation of personalized, social and environmentally aware (web and mobile) tourism services and territory monitoring services that leverage the regional networking and environmental infrastructures.

HARNESS

Homogeneous, commodity

cloud computing has

reached its limits

The dominant approach in offering cloud services today is based on homogeneous commodity resources: large numbers of inexpensive machines, interconnected by off-the-shelf networking equipment, supported by stock disk drives. However, cloud service providers are unable to use this platform to satisfy the requirements of many important and high-value classes of applications.

Today’s cloud platforms are missing out on the revolution in new hardware and network technologies for realising vastly richer computational, communication, and storage resources. Technologies such as field programmable gate arrays (FPGAs), general-purpose graphics processing units (GPGPUs), programmable network routers, and solid-state disk drives (SSDs) promise increased performance, reduced energy consumption, and lower cost profiles. However, their heterogeneity and complexity makes integrating them into the standard Platform as a Service (PaaS) framework a fundamental challenge.

The HARNESS project brings innovative and heterogeneous resources into cloud platforms through a rich programme of research, validated by commercial and open source case studies.

HARNESS integrates heterogeneous hardware and network technologies into data centre platforms, vastly increasing performance, reducing energy consumption, and lowering cost profiles for important and high-value cloud applications such as real-time business analytics and the geosciences.

AT A GLANCE

Project title:

Hardware- and Network-Enhanced Software Systems for Cloud Computing

Project reference:

Project coordinator:

Alexander L. Wolf, Imperial College London, UNITED KINGDOM

Partners:

École Poly. Fédérale de Lausanne, (CH) Université de Rennes I, (FR)

Zuse Institute Berlin, (DE) Maxeler Technologies, (UK) SAP AG, (DE)

Duration:

Total cost:

Website:

The HARNESS vision

The HARNESS project advances the state of the art in cloud data centre design so that: (1) cloud providers can profitably offer and manage the tenancy of specialised hardware and network technologies much as they do today’s commodity resources and (2) software engineers can seamlessly, flexibly, and cost-effectively integrate specialised hardware and network technologies into the design and execution of their cloud-hosted applications.

HARNESS develops an enhanced PaaS software stack that brings new degrees of freedom to cloud resource allocation and optimisation. A cloud application is seen to consist of components, some of which have multiple implementations. Applications express their computing needs to the cloud platform, as well as the price they are prepared to pay for various levels of service. This expression of needs and constraints builds upon what can be expressed through today's simple counts of virtual machines or amounts of storage, to encompass the specific and varied new factors characteristic of specialised hardware and network technologies.

The cloud platform will have access to a variety of resources to which it can map the components.

A flexible application may potentially be deployed in many different ways over these resources, each option having its own cost/performance/usage characteristics.

Specialised technologies are virtualised into resources that can be managed and accessed at the platform level. The idea is to provide flexibility to the platform as to which, when, and how many resources are used, and to separate that concern from the low-level deployment and monitoring of the concrete technology elements. Associated with the virtualised resources are policies that govern how allocation and optimisation decisions are made. Also associated with these resources are facilities to track their capacity, usage, and general availability.

Expected Impact

The public cloud services provider market is projected to reach nearly $22 billion by 2015. HARNESS will enable those providers to offer new levels of service to cloud applications at the same time as it opens a new market to the purveyors of specialised hardware and network technologies.

LEADS

Context of the project

The Web 2.0 revolution is transforming the Internet to a collaborative media where users can meet, read and write. The user-generated content constitutes a rapidly increasing proportion for the Web. Every day, 15 petabytes of new information is generated. Accessing and processing data in real-time will become more and more crucial in the near future to our data-driven society. The business of many companies is driven by online information (e.g. Web analytics, social networks, aggregators, search engines, etc.). Data growth is becoming the biggest challenge for enterprises to manage their own data centre hardware infrastructure. Clearly, the monetary investment required for crawling, storing, and processing even a small portion of the Internet is very high, making such a task intractable for small-scale and start-up companies. Currently, only the biggest information technology players have access to the infrastructure for storing huge amounts of data and the computing facilities needed to process it. Small and medium companies often have no other choice than relying on larger companies with dedicated data centres to provide them with the data and processing resources.

The monetary cost of the infrastructure is among the critical factors determining how to store big data. This problem is especially acute for small and medium companies that

The LEADS project will investigate a novel approach, based on a data-as-a-service model, such that the real-time processing of private and public data becomes economically and technically feasible even for small and medium enterprises. It will develop means to gather and store publicly available data, and will provide a platform that facilitates the real-time processing of this data. Moreover, the public data will be enriched with private data maintained and queried in a privacy-preserving manner on behalf of client applications. LEADS will operate on a collection of micro-clouds, independent and geographically distributed, and will be showcased by two industrial applications.

AT A GLANCE

Project title:

Large-Scale Elastic Architecture for Data-as-a-Service

Project reference:

Project coordinator:

Partners:

Technische Universität Dresden, (DE) Technical University of Crete, (GR)

Fundació Barcelona Media Universitat Pompeu Fabra, (ES)

Red Hat, (IR) AoTerra, (DE) Adidas, (DE)

Duration:

Total cost:

Website:

have limited resources. Therefore, any new solution should offer pricing competitive with, or lower than, conventional data centres to be attractive.

Data-as-a-Service

LEADS will provide an economical approach to process large amounts of data by sharing the collection, storage and querying of public and private data. LEADS will provide a DaaS framework that will permit users to gather and store public data, and enrich it with private data maintained on their behalf. Processing of real-time data can exploit the public and private data, including historical versions. The proposed DaaS framework will provide high efficiency and offer scalability as the demand grows.

Federation of micro-clouds

LEADS will investigate the design of a decentralized platform composed of a collection of micro clouds, each consisting of several servers. The micro-clouds will be based on technology provided by consortium member AoTerra and designed to reuse the waste heat of computing resources for other purposes, hence improving the energy efficiency of the whole architecture. Micro-clouds will be decentralized but will conceptually appear as a single cloud to clients. Geographic distribution will be exploited to provide faster access times, better utilization of network links, and improved availability and fault tolerance. By relying on a decentralized architecture, LEADS will solve the challenges of network

congestion and connectivity faced by cloud providers.

Impact of the project

LEADS results will have high potential for impact and results integration. Industrial partners have direct interest in exploiting the project outcome, as users or as service providers in a fast-changing information technology world. LEADS has the potential to significantly impact on a variety of software and hardware technologies, standards and open source, and on society as a whole by redefining how information is managed, collected and queried. All services using LEADS will benefit from the improved privacy and security. In turn, this will lead to an improved trust in these services.

LEADS will facilitate the design of new data-driven services. In particular, LEADS will simplify their development by providing an open interface. It will reduce the costs so that also SMEs can develop and operate data-driven services that would have previously been too expensive to develop and to operate.

MARKOS

Goals and objectives

Recent studies investigating the reuse of code in Open Source software projects show that developers in Open Source software projects commonly reuse available code and other knowledge that solves their technical problems.

Moreover, developers spend non-negligible amounts of time studying scientific publications and standard specifications, or learning from the source code (and its documentation) of related projects to reuse algorithms and methods without simply copying the code.

Therefore, developing a software system reusing existing Open Source solutions implies time consuming activities that are not performed when software is developed from scratch or without third party code. All this calls for the need to free the software analysts and developers from the technological barriers caused by the heterogeneity of approaches adopted by each Open Source project, by providing similar information on the software characteristics.

MARKOS intends to realize the prototype of an automatic service providing an integrated view on the Open Source projects available on the web, focusing on functional, structural and licences aspects of the software code released by the projects.

MARKOS realizes a prototype of an interactive application and a Linked Data API providing an integrated view on the Open Source projects available on the the web, focusing on functional, structural and licences aspects of software code. The MARKOS system itself will be released as Open Source software, which thanks to the offered functionalities, is expected to facilitate software development based on the Open Source paradigm in a global context.

AT A GLANCE

Project title:

The MARKet for Open Source - An Intelligent Virtual Open Source Marketplace.

Project reference:

Project coordinator:

Klaus-Peter Eckert, Fraunhofer-Gesellschaft e.V. Institute FOKUS, GERMANY

Partners:

Engineering Ingegneria Informatica S.p.A., (IT) ATOS SPAIN SA, (ES)

Poznan Supercomputing and Networking Center, (PL)

Geeknet Media, (UK)

Università degli Studi del Sannio, (IT) T6 Ecosystems, (IT)

Duration:

Total cost:

Website: