DIStributed Computing and InfOrmation Processing (DISCO) covers different research topics related with resource optimization, workload management, deployment, monitoring and operations of Future Internet infrastructures and platforms based on Cloud Computing paradigm.
The main topics covered by the current research activities include:
Highly available, distributed and automated cloud environments. Cloud paradigm is becoming more and more central in the ICT arena as solution to simplify the management of ICT resources and services. As such it is a fundamental building block of Future Internet research. DISCO explores architectures and tools for automated deployment, orchestration and monitoring of cloud resources from the IaaS layer to the PaaS layer ensuring high availability of the services (i.e. no single point of failure) and exploring different combination of vertical and horizontal elasticity (i.e. supporting flexible service load). Largely distributed cloud environments poses different challenges to current Internet technologies that DISCO wants to explore through real scale testbeds and real scenarios deployment. Our research activities in the field goes from traditional clouds, where computational capacities are hosted in data-centers, to edge clouds, where computational capacities are deployed closer to the users allowing for low latency applications.
Algorithms and models for the control and optimization of distributed and heterogeneous information and computing systems. The presence of flexible infrastructures for the deployment of applications on clouds and emerging processing paradigms requires to continuously match resources allocated in the system with respect to customers demands and the infrastructure operational state. Also, recent paradigms envision the federation of multiple clouds subject to diverse tenants and possibly heterogeneous business models. DISCO applies tools from optimal control, optimization and game theory in order to provide efficient resources allocation and pricing while preserving quality of service. Further areas of research are monitoring of nodes and applications done in order to trace cloud processes and the control of information diffusion processes for efficient content distribution.