MongoDB 與 MySQL Cluster 效能分析

本文探討MySQL Cluster和MongoDB的讀寫效能。MySQL Cluster是Oracle在MySQL資料庫下開發的一個叢集式架構的關聯式資料庫（RDBMS）；MongoDB是無關聯式資料庫（NoSQL）其中之一。兩種資料庫都提供了橫向擴展並且自動分割的機制，將原本單一台的資料量平均分散到其他機器上，讓讀取和寫入的負擔減少而加快效能，這次研究的主題除了兩種資料庫讀寫速度外，還會探討在橫向擴展後效能的提升幅度。在本論文中，實驗結果顯示在特定的資料格式與規模下 MongoDB 的讀寫效能較 MySQL Cluster 為佳，並反映節點的增加未能顯著改善資料庫系統之讀寫效能。

關鍵字

大數據； MongoDB ； MySQL Cluster ；無關聯資料庫；效能分析

並列摘要

This thesis focuses on the comparison of I/O performance between MySQL Cluster and MongoDB. MySQL Cluster is a cluster architecture that was developed by Oracle, and it is also a relational database（RDBMS）. Oppositely, MongoDB is one of the non-relational databases（NoSQL）. Both database systems provide a mechanism for horizontal expansion and can segment the database automatically. Thus, the database systems can disperse the data from a single machine to a cluster of machines. Therefore, speeding up the database I/O can improve the performance considerably. In this study, the analysis focuses on the I/O performance with different number of system nodes between the two databases. In this thesis, we show that with specific data format and scale, the I/O performance of MongoDB is better than that of MySQL Cluster.

並列關鍵字

Big Data ； MongoDB ； MySQL Cluster ； NoSQL ； Performance Analysis

參考文獻

[7]Maged Michael, Jose E. Moreira, Doron Shiloach and Robert W. Wisniewski. Scale-up x Scale-out: A Case Study using Nutch/Lucene. In proceedings of IEEE International Symposium on Parallel and Distributed Processing, 2007, pp. 1--8.

[1]陸嘉恒(2013)，挑戰大數據，臺北市。佳魁資訊。