September 18, 2009 at 7:11 am
Hi all.
I just need a little advice on database design, as I have been tasked to produce a schema to work with a new application.
Using my experience and a few article out there I am trying to designed a flexible and "performing" schema.
The application in question is of media/high transactions- 24/7. Over 200,000 records created daily. Each month we need to invoice these records without impacting on the current day-to-day use!
In this Scenario is a normalize or de-normalize schema the best. I know most of the answers will be "it depends", but Im trying to get an idea for what other people advise/think/know etc etc.
Thanks
JL
September 18, 2009 at 7:41 am
Data modeling 101: OLTP == 3NF 😉
_____________________________________
Pablo (Paul) Berzukov
Author of Understanding Database Administration available at Amazon and other bookstores.
Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.September 21, 2009 at 7:29 am
It does depend but if I had 200,000 inserts a day I would want to normalize the data as much as possible. After a month, your looking at 6,000,000 records and as long as it normalized it should hum. If it is not normalized you may see issues.
September 21, 2009 at 10:25 am
In your case, the answer is not "it depends." As tmacs33 said, normalization is best for you. For databases with heavy INSERTS and UPDATES, normalization will yield the best performance. For reporting databases, de-normalization will yield the best performance which is why a lot of organizations have their production db's normalized, and use a separate de-normalized datawarehouse for reporting.
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply