Advantages of dimensional models instead of relational mode

[system]

You could have a look at these posts for some ideas

120×120 10.4 KB

Hello, I am creating a small (<100,000 transactions) data warehouse for trending that requires many lookup tables for select criteria in queries. I’ve identified about 5 dimensions for the fact table, but my problem is that each transaction may require up to 20 different criteria (with lookup descriptions). The transaction dimension and the fact table both represent a specific transaction. My question is, how do I represent the 20 lookup descriptions per transaction? Should I keep this info in…

120×120 10.4 KB

Over 2 years ago, we paid a design team to develop universe for our site. I’ve recently acquired managing this mess and I’m wondering if any of you have the same issues: Users complain about not understanding the universe structure. I’ve posted the universe reports but the diagrams and list of joins/constraints don’t make sense to the average Joe user. The frustration comes in the ad-hoc arena where users pull objects that should be “joinable” and get errors or no data. Has anyone developed g…

120×120 10.4 KB

Hi All, I have a simple query, for which I require really a prudential and straight forward answer. My question is this: Data warehouse is meant to store historical information in it, and this is quite obvious. One of the main differences b/w STAR & SNOWFLAKE schema is said to reduce joins/links. This means, in snowflake environment we will have much normalised tables, which in turn consume less space comparitive to star. In an environment where we are storing historical data, what is the exac…

Plus I would recommend Ralph Kimballs book (available through Shop@BOB

---

irish_stan (BOB member since 2003-05-13)