DSpace Repository

Prise en compte de l’énergie dans la phase d’exploitation des bases de données volumineuses

Show simple item record

dc.contributor.author Amine ROUKH
dc.date.accessioned 2018-11-09T20:35:23Z
dc.date.available 2018-11-09T20:35:23Z
dc.date.issued 2017
dc.identifier.uri http://hdl.handle.net/123456789/884
dc.description.abstract In the Big Data Era, the management of energy consumption by servers and data centers has become a challenging issue for companies, institutions, and countries. In data-centric applications, Database Management Systems are one of the major energy consumers when executing complex queries involving very large databases. Moreover, the processing of this type of data requires costly and energy-intensive computing and hardware infrastructures. Current practices in the use and exploitation of very large databases indicate that the energy cost of query is totally neglected by users and also by designers. Knowing that the most important factor for the user is minimizing the response time of queries. In this thesis we propose a multi-objective formalization of the databases exploitation techniques, taking into account two non-functional requirements : performance and energy consumption during the execution of a queries workload. This formalization allow us to take advantage of the advanced techniques proposed in the state-of-the-art for solving multi-objective optimization problems. For this purpose, we rst develop cost models that estimate the energy consumption of queries executed in an isolated or parallel manner. These cost models are then integrated into one of the most important modules in a DBMS, which is the query processing module. The new objective of this module is the selection of execution plans of queries with the trade-o desired by the users between the time and the energy of the queries. Furthermore, we propose an initiative that integrates the energy dimension in the physical design phase of databases, in order to select optimization structures taking into account the energy aspects. We study the case of materialized views, one of the redundant optimization structures that is heavy used by database administrator. In each contribution of our thesis, intensive experiments are conducted using a real device for energy measurements and data of the TPC-H, TPC-DS and SBB benchmarks with various hardware and software con gurations. en_US
dc.language.iso other en_US
dc.subject Energy e ciency, cost models, query processing, physical design, energy management, multi-objective optimization. en_US
dc.title Prise en compte de l’énergie dans la phase d’exploitation des bases de données volumineuses en_US
dc.type Thesis en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search DSpace


Browse

My Account