Hello People,

I am currently in a phase of learning to create a Business Intelligence service and this is the project I am going to create as my final year project.

Let's suppose I have an access to the database of a company.Now, I want to create data marts or a datawarehouse from that database and treat it as an online server which can be accessed to apply business intelligence on it by the manager and get the results on his own client machine.

Remember, I am creating the data mart(s) or a datawarehouse on a separate machine(s) and the end-user or the manager is using his own local machine to get the BI service through the online connection with the server.

Now the questions arises here are as follows:

1.Should I create datawarehouse (i.e the integrated historical data store) or a set of datamarts (historical data stores for each individual department). What is the easiest and less time consuming approach to make this project work?

2.If my sole purpose is to make a good business intelligence service. Can i create a business intelligent service on a hypothetical data and not the real world data?

3.How big is a datawarehouse? I mean can I create a datawarehouse on a single or few machines if the database is not that large?

I will be waiting for replies...