Consolidating the Data

I hear a lot about hardware vendors coming out with consolidation services - server consolidation, data center consolidation, Cloud Computing etc. It has been on my mind that.. to truly get the advantage we need from this consolidation.. we will have to take one step back - to data consolidation.

The thing is, unless this data becomes comparable and consolidated, unless the database structure itself is consolidated, simplified, and quite simply, brought together, all we are doing is picking up tables and table and terabytes and terabytes from one server space to another server space. From 10 backup servers around the world to one main server set (Data Center) and one Disaster Recovery Data Center. But it doesn't matter where your data sits! Its still as much data!

Let me take an example. Data Center consolidation without data consolidation is like reducing the space occupied by the population of a village. Lets assume that the village has 100 people, who live in 20 houses. There is a need to consolidate the village. Therefore, 5 shelters are built and the entire village is asked to locate to these shelters - which are grouped together. 20 people per shelter. There is a scope to build another shelter if the population grows, so we have taken care of scalability. This saves physical space. But that's the ONLY thing that it saves. The food consumed by the people does not change, the other resources remain the same. they still need 100 pairs of clothing a day, etc. In addition, if one of the shelters is flooded or otherwise damaged, suddenly, 20 people are homeless. Because the space in other shelters is already compact, finding a home for these 20 people at one go is a task. (That is why we have disaster recovery backups, my dear Watson)

Am I against consolidation? ABSOLUTELY NOT. I am as pro consolidation as it can get. My point is that, while most hardware firms are focusing big time on hardware consolidation - data centers, servers etc., it is important to focus on the consolidation of that data itself. Removing redundancies. Using standard data and storing only deviations separately. Normalizing databases across company codes to the extent possible.

Everything that cannot be consolidated still remains separated, but we try and get all the data to talk to each other, to make sense relative to each other.. then, the data base structure itself will become simplified. And then, we will truly have "consolidated". Data Center consolidation is like population moving to community shelters. But real data consolidation is like stacking building blocks. If you try to store individual, separated building blocks in their original box, it never works out. They seem to have magically grown to too many. But if you put them together in a design, its amazing how little space they take. Furthermore, each connection makes SENSE.

THAT is the difference that one is talking about.

This article is contributed by Nidhi Arora, Co-Founder and Director at Topgain Consulting Pvt. Ltd.