Nowadays we are witnessing the diffusion of the Function-as-a-service paradigm, that has emerged due to its ability to deploy fine-grained units of computation on heterogeneous resources on-demand, reducing costs and improving resource utilization. In this work, we discuss our objective: extending the runtime support of a structured parallel programming framework, FastFlow, to empower it with the mechanisms needed to exploit FaaS functionalities. To achieve this objective, we also integrates workflows as a way to specify and describe, at a very high level of abstraction, the topology and the requirements of complex parallel data stream applications.
StarFlow: fine-grained execution of workflows in Hybrid Cloud HPC for data stream applications
Ferrucci L.;Danelutto M.;Dazzi P.
2023-01-01
Abstract
Nowadays we are witnessing the diffusion of the Function-as-a-service paradigm, that has emerged due to its ability to deploy fine-grained units of computation on heterogeneous resources on-demand, reducing costs and improving resource utilization. In this work, we discuss our objective: extending the runtime support of a structured parallel programming framework, FastFlow, to empower it with the mechanisms needed to exploit FaaS functionalities. To achieve this objective, we also integrates workflows as a way to specify and describe, at a very high level of abstraction, the topology and the requirements of complex parallel data stream applications.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
