Private cloud infrastructures are now widely deployed and adopted across technology industries and research institutions. Although cloud computing has emerged as a reality, it is now known that a single cloud provider cannot fully satisfy complex user requirements. This has resulted in a growing interest in developing hybrid cloud solutions that bind together distinct and heterogeneous cloud infrastructures. In this paper we describe the orchestration approach for heterogeneous clouds that has been implemented and used within the INDIGO-DataCloud project. This orchestration model uses existing open-source software like OpenStack and leverages the OASIS Topology and Specification for Cloud Applications (TOSCA) open standard as the modeling language. Our approach uses virtual machines and Docker containers in an homogeneous and transparent way providing consistent application deployment for the users. This approach is illustrated by means of two different use cases in different scientific communities, implemented using the INDIGO-DataCloud solutions

In this paper we propose a distributed architecture to provide machine learning practitioners with a set of tools and cloud services that cover the whole machine learning development cycle: ranging from the models creation, training, validation and testing to the models serving as a service, sharing and publication. In such respect, the DEEP-Hybrid-DataCloud framework allows transparent access to existing e-Infrastructures, effectively exploiting distributed resources for the most compute-intensive tasks coming from the machine learning development cycle. Moreover, it provides scientists with a set of Cloud-oriented services to make their models publicly available, by adopting a serverless architecture and a DevOps approach, allowing an easy share, publish and deploy of the developed models. INDEX TERMS Cloud computing, computers and information processing, deep learning, distributed computing, machine learning, serverless architectures.

As different Cloud Management Frameworks and resource providers are settling in the market there is the need to grasp the interoperability problems present in current infrastructures, and study how existing and emerging standards could enhance user experience in the cloud ecosystem.In this paper we will review the current open challenges in Infrastructure as a Service cloud interoperability and federation, as well as point to the potential standards that should alleviate these problems.

SummaryCloud computing has permeated into the IT industry in the last few years, and it is nowadays emerging in scientific environments. Science user communities are demanding a broad range of computing power to satisfy high-performance applications needs, such as local clusters, High Performance Computing (HPC) systems and computing grids. Different workloads need from different computational models, and the cloud is already considered as a promising paradigm.The scheduling and allocation of resources is always a challenging matter in any form of computation and clouds are not an exception. Science applications have unique features that differentiate their workloads, hence their requirements have to be taken into consideration to be fulfilled when building a Science Cloud. This paper will discuss what are the main scheduling and resource allocation challenges for any Infrastructure as a Service IaaS provider supporting scientific applications.