Please correct me if i am wrong.
1.Instead of using a star schema to report off, can i create a de-normalized schema and have constraints taken care at a reporting level?? I know this is not ideal, but i am the not having the resources for a data modeler and neither have erwin or the higher
grade ETL tools. Needless to say, the shear amount of time and resources it would take to test the schema if i add constraints at the schema level and the errors that i would encounter during Load. I need a build, i'm ok with a dirty build, but i need to get
All i have is Std. SQL Server 2008. I want to incorporate SSIS, and SSAS to provide the reporting teams with a dw solution. Thanks.
2. Please provide the step. Please correct me if i am wrong.
a. Denormalize OLTP schema (relevant tables) and bring the data together = staging
b. apply Business rules and transformations.
c.Load into de-normalized tables = similar to data marts
d.convert to multidimentions for analysis....cubes. (How do i choose/decide measures(facts?) and how do i decide dimentions???) ( is time a fact?? in most cases? is this always true? if not please explain with example)
View Complete Post