Resolved Need a Database Admin/Designer to interview for my ITD Class

Hello all,

I know this is random and I may be in the wrong forum, but I'm not sure where else to turn (I've always had prompt and awesome responses from this community.

My database teacher and textbook assume I have easy access to a database administrator/designer to ask a couple questions about normalization and denormalization and submit their responses for a small part of my grade. Of course, I know no one in this field, so I'm hoping there may be someone here that I can ask two quick questions to? Or if you know someone that you may be able to ask on my behalf and act as a middleman, that would be acceptable too.

Feel free to reply here or send me a direct message and I'll get the two questions to you ASAP.

Thank you (hopefully) in advance!

 

What database are you using and what do you need to know?

 

It's just an intro to database that's part of my IT even though I'm going in networking. We're just covering normalization and denormalization of database design and implementation (using relations if it makes a difference).

The first question is whether you normalize your relations higher than 3rd Normal Form and why you do or don't go beyond 3NF.

Second is whether you believe in fully normalized or denormalized physical databases and why.

 

Its difficult to give a concise answer to these questions in a forum like this. To answer these questions fully would take a book!!

However, I'll try and answer in a couple of sentences.

Normalization is the avoidance of the duplication of data in a relational database. In essence, its the difference between flat file and relational databases. You should check this page for a short explanation of Normal Forms.

Like most things in computing, these things are a "trade off" between minimising the bloat in the database created by storing multiple instances of the same data and the load placed on the database server by the query. Normalization breaks the data up into multiple unique chunks with the consequential load on the server when a query is run because the server has to then look in multiple locations to put the output together. This is essentially why you don't go beyond 3NF. The benefits of normalization are neutralized by the excessive load placed on the server.

This basically answers your second question too as both questions are essentially the same question but from different directions. In a denormalized database, the reverse is the case as the benefits of minimal server load are neutralized by the bloat that is created by duplicating data multiple times.

Does that make sense?

 

Just came across this page which will give you some useful information.

 

Thank you Macleod; I already knew all about normalizing and denormalizing, we just needed an actual industry professional's opinion on it, so I appreciate the response =)

 

You're welcome. When I did my Microsoft Certification in SQL Server, if I remember correctly, we did the best part of a day on it. Its difficult to summarize Normalization in a few sentences especially when I don't know how familiar you are with the terms used.