| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 455.35 KB | Adobe PDF |
Advisor(s)
Abstract(s)
As the amount of information flow increases, business companies feel the need to improve on storage systems. Henceforth, to tackle this increasing need, paradigms such as NoSQL emerge to solve the unlimited data growing requirement. However, the NoSQL solution has no proofs given in the field to support their solution claims. Benchmarks can test and compare different solutions performance by executing queries over a toy dataset (synthetically generated). The problem with benchmarking results is how to extend the conclusions to a real system operating within a real business scenario. In this paper, an actual corporate case study is used, with real-world data, to evaluate how NoSQL databases perform. First, using big data, write-intensive tests are implemented and evaluated using Cassandra, MongoDB, Couchbase, and compared with the relational database in place, which is within the throughput limit. Results show a throughput comparison for each tested approach.
Description
Keywords
Big-data NoSQL database ETL
Citation
Publisher
Springer International Publishing