By Petr Hosek, www.linkedin.com/in/phosek
We live in times when bold and aggressive statements bear more weight than real deeds. This is also valid for the Big Data world. We hear loud
proclamations from players - who just encapsulated a bit of open source code and brought it to the market – how they will save the world!
The trouble is – the majority of these new tools serve usually one purpose: Fast data only. Or web-clicks only. Or Social posts only. Also, most of the new
types of data are of low density – that means the ratio between the valuable information and the physical amount of data it is hidden in is rather low.
However, let us take a step back: Where is the most valuable information stored today? It is clearly in relational database systems! With data typically
created through enterprise applications, the information density of RDBMS is incomparably higher!
When you discuss today’s major challenges in data and information management, one of the top tickets is data silos! And creating
additional ones for big data (social posts; sensory data; location data, etc.) sounds like the worst nightmare to many CIOs!
As a result, the holy grail of Big Data is therefore not to work with the new sorts of data in isolation, but to connect the unstructured or
semi-structured data with the existing relational data. In other words, the game is not about Big Data, it is about All Data!
We will for sure see in future a few distinctive data platforms that are tuned for certain type of data. As of today, three of them are clear leaders.
RDBMS with its robustness and functionality will doubtlessly sustain its central role in Information management. And we will see more and more data stored
on NoSQL and Hadoop stacks. Hence, a typical corporation will entertain a combination of different platforms for a cost-effective store, and will use tools
that allow to connect, combine and to correlate all these data no matter where they reside.
technology will merge the two traditional data worlds of today, Online Transaction Processing and Data Warehousing. And we will see more and more tools
that encompass the all above mentioned platforms; Oracle Big Data SQL being an example.
Let me finish with a small hint: Whenever you think about Big Data, even if you want to start with a small isolated pilot, always think about a long-term
picture and that undoubtedly means about All Data!
To learn more about Oracle’s approach to All Data Information management, read this white paper.