Alocação de dados em bancos de dados distribuídos

Matheus WildembergMelise M. V. PaulaFernanda BaiãoMarta Mattoso

The problem of data allocation directly impacts on the execution cost of the application over a distributed database in environments that work with data distribution and replication, such as web servers, heterogeneous database integration systems and distributed databases. Allocation algorithms are typically used to find a data distribution among the sites of the network such as to minimize the execution cost of the application. In this work, a heuristic algorithm is proposed for fragment allocation in distributed database systems. The proposed algorithm is based on a heuristic algorithm presented in the literature, and its goal is to find out an allocation schema with a minimal execution cost, maintaining the complexity of the original algorithm. Through simulations performed on top of the TPC-C benchmark, it was possible to identify scenarios where the proposed algorithm found the optimal solution and other scenarios where the proposed algorithm found an allocation schema with a reduced cost when compared to the original algorithm.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: