17
p. 1 1 Chapter 2 - Video # 6 Where Should I Put My Data Warehouse? 1 Chapter 2: Business Intelligence & Data Warehousing with SSAS Course: SQL Server 2008/R2 Analysis Services Course Id: 165 Presented by Scott Whigham

Where Should I Put MyData Warehouse?

Embed Size (px)

DESCRIPTION

Where Should I Put My Data Warehouse?

Citation preview

Page 1: Where Should I Put MyData Warehouse?

p. 1

1

Chapter 2 - Video # 6

Where Should I Put My Data Warehouse?

1

Chapter 2: Business Intelligence & Data Warehousing with SSASCourse: SQL Server 2008/R2 Analysis ServicesCourse Id: 165Presented by Scott Whigham

Page 2: Where Should I Put MyData Warehouse?

p. 2

2

• Overview of Chapter

• Defining Business Intelligence

• BI and SQL Server

• OLTP vs. OLAP

• “Where should I put my data warehouse?”

• Multi-dimensional databases

• Data Mining

• What is Analysis Services?

• New Features in SSAS 2008

Business Intelligence

Page 3: Where Should I Put MyData Warehouse?

p. 3

3

• So we have a few basic terms:– Data warehouse

– Data mart

– OLTP database

– OLAP database

Where, oh where…

Page 4: Where Should I Put MyData Warehouse?

p. 4

4

Where, oh where…

• We even have a quasi-fancy diagram:

Page 5: Where Should I Put MyData Warehouse?

p. 5

5

• What we are missing is exactly who goes where and why– Not an easy answer

– Cannot simply say, “Analysis Services stores the OLAP and SQL Server handles OLTP”

Where, oh where…

Page 6: Where Should I Put MyData Warehouse?

p. 6

6

• Large OLTP databases reside on relational database management systems (RDBMS)– SQL Server, Oracle, MySQL, DB2

– OLTP databases cannot reside on Analysis Services

Where, oh where…

Page 7: Where Should I Put MyData Warehouse?

p. 7

7

• OLAP databases may reside on either an RDBMS or a dedicated data warehousing product like Analysis Services

Where, oh where…

Page 8: Where Should I Put MyData Warehouse?

p. 8

8

• OLAP simply defines how we store/use data for analytical purposes– It does not define the structure of the data; only

that the structure should make for quick and easy retrieval

– OLAP can be relational or multidimensional

• We’ll cover “multidimensional” in the next video

Where, oh where…

Page 9: Where Should I Put MyData Warehouse?

p. 9

9

• Relational data warehouse– OLAP database stored on an RDBMS server

– Uses a set of dimension and fact tables

• Discussed in a future video in this chapter…

– Report server can access data directly for saved and/or ad-hoc reporting

Where, oh where…

Page 10: Where Should I Put MyData Warehouse?

p. 10

10

• Relational data warehouse (cont.)

– Must decide if you are using DW for direct analysis or for archival reasons

• An archival system will store records on a 1:1 basis– Copies/archives records from data entry systems

• Aggregations are not pre-calculated

• Generally speaking, a relational data warehouse will either store aggregates or archival but not both

Where, oh where…

Page 11: Where Should I Put MyData Warehouse?

p. 11

11

• Relational data warehouse (cont.)

– Example of a relational data warehouse storing product data

Where, oh where…

Page 12: Where Should I Put MyData Warehouse?

p. 12

12

• Relational data warehouse (cont.)

– Example of a relational data warehouse storing archival data

Where, oh where…

Page 13: Where Should I Put MyData Warehouse?

p. 13

13

• Multidimensional data warehouse– OLAP database not stored on an RDBMS but on a

multidimensional database management system

– Uses cubes, measures, and dimensions

• Discussed in a future video in this chapter…

– Report server can access data directly for saved and/or ad-hoc reporting

Where, oh where…

Page 14: Where Should I Put MyData Warehouse?

p. 14

14

• Multidimensional data warehouse (cont.)

– Strictly for analysis, not archival

– Stores only pre-calculated aggregates

• Not detail records

– Analysis Services is the multidimensional database server for Microsoft!

Where, oh where…

Page 15: Where Should I Put MyData Warehouse?

p. 15

15

• So to answer the question, “Where should I put my data warehouse?”– The answer is likely to be “You need both a

relational data warehouse to store your archival records and a multidimensional data warehouse to store your pre-calculated aggregates!”

Where, oh where…

Page 16: Where Should I Put MyData Warehouse?

p. 16

16

Where, oh where…

• Let’s expand our diagram:

Page 17: Where Should I Put MyData Warehouse?

p. 17

17

• Multidimensional Databases and Datamarts

“Human beings were not meant to sit in little cubicles staring at computer screens all day, filling out useless forms and listening to eight different bosses drone on about mission statements.”

Peter Gibbons, Office Space

Next up