The Curious Case of Database Deduplication

In the recent years, deduplication technologies have become increasingly popular in the modern datacenter. Whether it be purpose built backup appliances that use deduplication to reduce the footprint of backups, or be the all flash arrays that utilize deduplication to get better storage efficiency out of flash storage. Though all deduplication technologies are not created equal but they do promise significant reduction in footprint of the data actually stored on media. The amount of data reduction not only depends on the deduplication technology but also the type of data that is being duplicated. While majority of the storage appliances advertise huge storage savings when it comes file system data, the duplication ratios for data stored in relational databases is significantly lower. This session makes an attempt is analyzing the reasons behind low deduplication ratios for relational databases and also contrast and compares various deduplication techniques in the context of relational databases.

Learning Objectives

Acquire a deeper technical knowledge of the deduplication technologies available
Understand the implications of using these technologies on relational databases
Develop a comparison framework to assess the suitability of various deduplication