Tying Together Olap, Datawarehouses, Datamarts, and 3-Tier Architecture

Write a short paper explaining OLAP, Data Warehouse and Data Mart, and Three-tier architecture.

OLAP

I read several articles before I completely understood what OLAP really meant. First, I found a definition that sounded fairly simple at this address on the web, http://www.bitpipe.com/tlist/OLAP.html. "Database software that provides an interface so users can transform or limit raw data quickly and interactively examine the results in various dimensions of the data". But then it went on to say "Also called: Online Analytical Processing, OLAP Cube, On-line Analytical Processing Software, Multidimensional Online Analytical Processing, MOLAP, Multidimensional OLAP, and Online Analytical Processing Software". Well this was not simple! I assumed this to mean that OLAP could be defined in many different ways, by many different vendors. So I read a couple of White Papers, perused a few others and realized they were all trying to sell their OLAP tools. Then I decided to go to my favorite search engine, Google. I found The OLAP Report site, which seemed to agree with my assessment of OLAP in the first place; that OLAP could mean something different to whoever was defining it. Who should you listen to? The OLAP Report Site gave a great test for a system claiming to be OLAP. It can be remembered in an easy to remember acronym, FASMI, or Fast Analysis of Shared Multidimensional Information. Here is their definition of FASMI.

FAST means that the system is targeted to deliver most responses to users within about five seconds, with the simplest analyses taking no more than one second and very few taking more than 20 seconds.

ANALYSIS - targeted for the user, this analysis should be able to cope with the business logic, statistical analysis used in the application, and should be kept simple.

SHARED - All security requirements should be done by the system, even down to the cell level. Concurrent update locking, at an appropriate level, should be used if multiple write access is used.

MULTIDIMENSIONAL - this is the key requirement in the FASMI test. The system must provide a multidimensional conceptual view of the data, including full support for hierarchies and multiple hierarchies.

INFORMATION is all of the data and derived information needed, wherever it is and however much is relevant for the application.

The FASMI test seems to be an understandable definition of the goals OLAP is meant to achieve.

Data Warehouse and Data Mart

To be competitive today, a business has to consolidate and aggregate data, so that they can leverage their information. They must take their data from production systems and place it in a centralized data warehouse or data mart for users to use. The can use that data for better customer service, do analysis for reporting or problem solving. Another reason to use Data Warehouses and Data Marts is for data mining. A Data Mart can be used at either end of the data mining process.

What is the difference between Data Warehousing and Data Marts? A Data Mart is subject or department -oriented. They can be subsets of larger warehouses, which make them dependent data marts. But they don't always have to be part of a data warehouse. They can be independent data marts called 'stovepipe' data marts. Size is not an issue when you are comparing a data mart and a data warehouse. Some data marts are larger than some data warehouses. They are similar and they can contain all of an organization's data. But a data mart can be more limited in scope. Byte.com states that "It will typically focus on the needs of a specific business unit or function and is less expensive and faster to implement