Cloud computing has generally involved the use of specialist data centres to support computation and data storage at a central site (or a limited number of sites). The motivation for this has come from the need to provide economies of scale (and subsequent reduction in cost) for supporting large scale computation for multiple user applications over (generally) a shared, multi-tenancy infrastructure. The use of such infrastructures requires moving data to a central location, undertaking processing on the data, and subsequently enabling users to download results of analysis. However, when applications need data to be captured and processed/analysed in real time, migrating all the data to a central site prior to analysis can create significant overhead. Examples include sensor network-based applications, such as smart buildings and smart grids, where sensors interface with real world artefacts and must respond to physical phenomenon that cannot be predicted apriori. The amount of data likely to be generated by a sensor and processing requirements in such applications cannot be pre-determined - they are often dependent on the rate of change of the physical phenomenon being measured and potential occurrence of “trigger events” which are non-deterministic.
Recently, there has been significant interest in creating “multi-clouds” or Cloud-of-Clouds federation to aggregate capabilities and capacities offered by a variety of cloud providers. Some of the efforts are focused on cloud interoperability. For example, the Open Cloud Computing Interface (OCCI) effort (OCCI), which defines a common interface for cloud providers; and the European FP7 “UNIFY” project (UNIFY EU FP7 project), which develops a Cloud Operating System (CloudOS) to connect distributed clouds and make use of in-network capabilities to process data (GENICloud project (GENICloud project)). Similarly, on-line sites such as CloudHarmony (CloudHarmony) report over 100+ cloud providers that offer capability ranging from storage and computation to complete application containers that can be acquired at a price, primarily using service-based access models. From the user’s perspective these environments bring a variety of benefits: (i) reduced reliance on a single vendor’s infrastructure; (ii) improved fault tolerance, as failure in one cloud system does not render the entire infrastructure inoperable; (iii) improved security – similar argument to fault tolerance, i.e. a breach in one cloud system does not impact the entire infrastructure; and (iv) the ability to utilise capability (and data) that may only be available in one cloud system and may not be transferable due to volume or legal constraints.