STARCloud · Bringing Quality of Service to Blockchain-Based Clouds

Today the cloud market is restricted to few providers, which entails higher prices, vendor lock-in and lower service quality. Smart Contracts (SC) and blockchain could reshape this market since they lower costs and provide trust without resorting to intermediaries. This paradigm, named blockchain-based clouds, prevents vendor lock-in and has the potential to make market's barriers so low that anyone with idle or spare computational resources can offer cloud services. The STARCloud project systematically utilises formal methods for the static analysis and runtime control of the next generation clouds for creating support for Quality of Service (QoS). To this aim, since the state-of-the-art solutions cannot be used in this domain, this project provides novel approaches and tools, including consensus mechanisms, a domain specific language for QoS definitions and QoS formal analysis tools, and a framework for the matchmaking and negotiation of SCs. Supporting QoS will boost the adoption of decentralised clouds; thus reducing prices and environmental impact; and improving QoS, which make the cloud providers and consumers that use blockchain-based clouds more competitive, specially in Europe, where the most advanced commercial solutions are located. Moreover, these solutions will enable geographically sparse micro and traditional cloud providers to join the market, which will leverage cloud and edge computing in Europe. The project is timely and, besides involving collaborators from many of the best European research institutes, it is supervised by I. Brandic, one of the main cloud experts in the world and actively involves one of the main players in the market: iExec. Overall, the solutions proposed here will contribute to the disruptive blockchain-based cloud paradigm and find immediate use in industry and academia by addressing the fundamental requirement of most cloud consumers, in particular business: QoS definition and enforcement in completely distributed systems.