Análisis comparativo del desempeño computacional entre dos gestores de bases de datos en consultas sobre la base de datos UniProt/Swiss-Prot

Abstract

The large biological data sets to be generated in the immediate future as a result of the usage of new sequencing technologies requires high performance applications to manipulate databases. Although, the storage costs of the worldwide have decreased, the availability and computational cost required to process the data is still growing. Thus, it is necessary a higher optimization in the database manipulation. Being this way how bioinformatics solves the problem of the organization of information and manipulation of large data sets. This paper proposes to test which NoSQL database manager, Elasticsearch or Hadoop, provides a better computational performance in terms of execution time for CRUD operations in data of proteics database UniProt/Swiss-Prot. The criteria used to identify which database manager is the more efficient tool in computational cost related to time is the delay of execution of several queries individually evaluated. The CRUD (Create, Read, Update, Delete) operations will be sent from scripts as project benchmarks through web services to each database manager on UniProt/Swiss-Prot database. The comparative analysis will be executed in a cluster environment with three nodes for each manager tool. The comparative analysis has been realized in a cluster environment with three nodes for each management tool. These nodes will be virtual machines with similar characteristics like same memory ram, storage and operative system.