Data Modeling for Cassandra

Authors

  • Thiago B Leão Exército Brasileiro, Instituto Militar de Engenharia. Rio de Janeiro, RJ, Brasil.
  • Maria Claudia R Cavalcanti Exército Brasileiro, Instituto Militar de Engenharia. Rio de Janeiro, RJ, Brasil.
  • Raquel C G Pinto Exército Brasileiro, Instituto Militar de Engenharia. Rio de Janeiro, RJ, Brasil.

Keywords:

data modeling, NOSQL, materialized views

Abstract

Due to the increasing demand for scalability and distribution of data, a variety of NOSQL database management systems have emerged and are being used for different purposes. Some of the most popular systems are: Document-oriented, Column-oriented and Key-value. Although there are many of these systems, to the best of our knowledge, there are barely any guidelines for data modeling for them. The proposed algorithm analyzes a predefined set of queries and, based on their filter clauses, it defines the composition of the primary and clustering keys for a set of materialized views. In addition, it defines a set of <query, materialized view> pairs, indicating which queries each materialized view addresses. To evaluate the algorithm, we report on an experiment that compares the performance of having different tables for each query and of having the materialized views suggested by the algorithm. It shows promising results and points to new directions on data modeling for NOSQL systems.

Downloads

Download data is not yet available.

Published

2022-01-19

How to Cite

Leão, T. B., Cavalcanti, M. C. R., & Pinto, R. C. G. (2022). Data Modeling for Cassandra. Revista Militar De Ciência E Tecnologia, 38(4), 70–82. Retrieved from https://ebrevistas.eb.mil.br/index.php/CT/article/view/9043