Feature Stories - Seeding clouds among European SMEs Europe's SMEs now have a reference architecture for deploying their own next-generation cloud computing thanks to the work of a multinational team of EU-funded researchers. It's a chance for small companies to start thinking really big. Digital Economy © Shutterstock Cloud computing refers to information technology - software, data and even processing power - that is accessed via a network. The business model works like a utility, where users can get as much electricity or water as they want, but only pay for the 'distributed computing' resources they use. The market for cloud computing will be worth EUR 13 billion in 2014, up from EUR 3 billion in 2010, according to a January report by research firm In-Stat, while Gartner, another research company, believes cloud computing will be as influential as e-business. Clouds are also of strategic importance to Europe and its ambitious Digital Agenda. Indeed, European Commission Vice-President Neelie Kroes, who is responsible for the Digital Agenda, plans to deliver a Cloud Computing Strategy for Europe by the end of 2012. So far, Europe has played a small role in the development of cloud computing infrastructures, but that is changing dramatically thanks to the work of the 'Resources and services virtualisation without barriers' (Reservoir) project funded by the EU's Seventh Framework Programme. It has created advanced open source software and tools to let small and medium-sized companies develop their own cloud platforms. 'Today's clouds are limited by the hardware behind them which, in turn, is typically limited by the size of a data centre hosting the cloud,' explains Eliot Salant, coordinator of the Reservoir project, and researcher with IBM's Haifa Lab in Israel. 'This is a premise of Reservoir - no single provider, however large, can meet the exponentially growing demand for resources.' Reservoir was set up to overcome this limitation. The software developed by the project allows cloud providers to federate, or combine their resources. If a cloud provider starts running out of resources, it can rent resources from another cloud provider. Clouds are a major trend in computing. The idea is that companies that want to offer services on the internet do not need to pay up front for large investments in equipment (CAPEX). They do not pay for the operating expenditure (OPEX) either. Instead, they pay a cloud provider on a per-use basis for hosting their service. This creates greater control of expenditure and avoids costly asset acquisition and maintenance. Hosting a cloud computing infrastructure was a logical business model for companies like Amazon and Google to adopt; these companies had huge, expensive data centres to run their own businesses, building up enormous data centre management expertise in the process. Offering to rent spare capacity in their server farms capitalised on their investment and leveraged their expertise. These hosting services have now become so successful Amazon could one day earn as much from cloud provision as it does from online shopping. But while US-based companies have been quick to offer cloud hosting, European firms have made less of an impact. In part, this is a question of scale. Most of Europe's technology industry revolves around small and medium-sized enterprises, while in the US giant corporations predominate. As such, fewer IT companies have the resources to offer cloud services on the scale available to their US counterparts. Scale up, think big But technology developed by the Reservoir project helps to solve this problem. The project sought to enable large-scale deployment and management of complex IT services both within and across clouds. The project attracted 13 leading industrial and academic partners from Europe and Israel and reads like a who's who of data services with heavyweights like IBM, Germany's SAP, Thales in France, Spain's Telefónica, and Oracle. The project not only sought to provide a platform that lets small data centre owners combine their resources into a federated cloud, it also sought to address many of the shortcomings of the current technology. For example, most cloud hosting technology is closed, proprietary software that is incompatible with other platforms, so Reservoir worked to define open standards for cloud computing, to break the lock-in imposed by many vendors today. It targeted cloud computing based on quality of service (QoS), backed by enforceable service level agreements, or SLAs, including the necessary monitoring tools at all administration levels. It researched techniques for rapid deployment of an application without service provider intervention, which will make cloud services more flexible and adaptable. The team also worked on automatic life-cycle management of services, again simplifying cloud management. Reservoir demonstrated how virtual resources could be seamlessly migrated across a cloud, which is a bit like carrying a performing musician from one room to another without dropping a note. It showed live migration of virtual machines from one physical host to another, cutting across network boundaries simply and easily. Additional innovations added security to the Reservoir environment, a rising concern as this emerging industry becomes larger and more pervasive. Towards commercialisation Much of the technology is also open source. 'We talk about the Reservoir framework, which is a collection of all artefacts that the project has produced,' says Mr Salant. 'This includes open source code, documentation, and proprietary code which can be licensed. Not all of Reservoir is open source and therefore you cannot just download and install it.' Nonetheless, the project has taken steps to ensure there are the fewest possible barriers to exploiting its work. 'The portions which are proprietary are described, however, in documentation which is publically available as part of the framework,' Mr Salant explains, 'Therefore organisations wishing to build their own Reservoir-compatible stack can either [do it] based on the specifications, or take the open source parts of the Reservoir framework and either license the proprietary parts, or build their own versions of them.' Finally, Reservoir had a high impact in the field, delivering 64 scientific publications and over 100 project presentations, keynote lectures, workshops and demonstrations. The project also secured 120 articles and news clips in the press which demonstrate the wider interest in the research and this field. 'Besides the value that came directly out of the project's research, I believe that the impact of Reservoir will continue to be felt - partners are working on commercialising spin-off technologies, a new generation of European cloud researchers has been trained, and follow-on research projects under the FP7 funding are going on,' notes Mr Salant. Indeed, Reservoir has seen a large number of spin-out technologies emerge from the project. For example, one spin-off from Reservoir was the so-called Lattice framework, developed by University College London, which is a monitoring tool to oversee large-scale, highly-dynamic IT environments. In many respects, with the work done by Reservoir, cloud computing is reaching its maturity. The promise of the cloud has always been that users did not have to worry about where the data, application or processing power was originated. Reservoir has pushed that paradigm even further, allowing resources to be seamlessly obtained across a federation of clouds. The Reservoir project received EUR 10.53 million (of the EUR 17.17 million total budget) research funding under the EU's Seventh Framework Programme, 'Service and software architectures, infrastructures and engineering' sub-programme. Useful links: - 'Resources and services virtualisation without barriers' - Reservoir project data record on CORDIS - Digital Agenda for Europe Related articles: - Cloud computing research precipitates spin-off tech - EU project to smooth the path through data-intensive environments - Architects and engineers bridge the grid chasm - Good view from the clouds
