Cloud Computing has made possible flexible resources provisioning from an almost unlimited pool. This has created the opportunity to broaden the horizon of data that can be analyzed, allowing to support the so called Big Data Analytics applications. New programming paradigms, such as NoSQL queries and Map-Reduce applications, have emerged within frameworks such as Microsoft Azure, Hadoop and Apache Spark. In many cases, applications execute jobs that are split into stages, each one composed of tasks that can be run in parallel on many computational nodes. Directed acyclic graphs describe the precedence between stages, defining the execution rules and controlling the degree of parallelism. This work presents a Process Algebra dialect aimed at describing both jobs and execution environments. The proposed framework is then used to model and study standard parallel programming benchmarks, to demonstrate its applicability.

Barbierato, E., Gribaudo, M., Iacono, M., Map-reduce process algebra: a formalism to describe directed acyclic graph task-based jobs in parallel environments, in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), (Moscow, 21-25 October 2019), Springer Science and Business Media Deutschland GmbH, Basilea 2020:12023 85-99. [10.1007/978-3-030-62885-7_7] [http://hdl.handle.net/10807/202863]

Map-reduce process algebra: a formalism to describe directed acyclic graph task-based jobs in parallel environments

Barbierato, E.;
2020

Abstract

Cloud Computing has made possible flexible resources provisioning from an almost unlimited pool. This has created the opportunity to broaden the horizon of data that can be analyzed, allowing to support the so called Big Data Analytics applications. New programming paradigms, such as NoSQL queries and Map-Reduce applications, have emerged within frameworks such as Microsoft Azure, Hadoop and Apache Spark. In many cases, applications execute jobs that are split into stages, each one composed of tasks that can be run in parallel on many computational nodes. Directed acyclic graphs describe the precedence between stages, defining the execution rules and controlling the degree of parallelism. This work presents a Process Algebra dialect aimed at describing both jobs and execution environments. The proposed framework is then used to model and study standard parallel programming benchmarks, to demonstrate its applicability.
Inglese
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
25th International Conference on Analytical and Stochastic Modelling Techniques and Applications, ASMTA 2019
Moscow
21-ott-2019
25-ott-2019
978-303062884-0
Springer Science and Business Media Deutschland GmbH
Barbierato, E., Gribaudo, M., Iacono, M., Map-reduce process algebra: a formalism to describe directed acyclic graph task-based jobs in parallel environments, in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), (Moscow, 21-25 October 2019), Springer Science and Business Media Deutschland GmbH, Basilea 2020:12023 85-99. [10.1007/978-3-030-62885-7_7] [http://hdl.handle.net/10807/202863]
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10807/202863
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact