Dimensional Modeling in Oracle SQL Developer Data Modeler

1

Copyright © 2013, Oracle and/or its affiliates. All rights reserved.

Dimensional Modeling in Oracle
SQL Developer
Data
Modeler
Philip Stoyanov
Senior Software Development Manager

2


Legal
The following is intended to outline our general product direction. It is intended for
information purposes only, and may not be incorporated into any contract. It is not a
commitment to deliver any material, code, or functionality, and should not be relied
upon in making purchasing decisions. The development, release, and timing of any
features or functionality described for Oracle‟s products remains at the sole
discretion of Oracle.

3


Agenda



Reporting



4

Dimensional model

Dimensions in physical model for Oracle


Models
Data Types
Model

Subject
Area 1

Subject
Area n

Relational
Model 1
Process Model
Data flow
diagramming

Logical
Model

5

…

Subject
Area 1


Dimensional
Models 1..n

Subject
Area n

…

…
Relational
Model n

DB Site 1
Oracle 9i

…
DB Site n
Oracle 12c

DB Site 1
DB2/390 v.8

…
DB Site n
MS SQLS 2005

Relational model without relationships
 Even if we add

foreign keys still
there are
dependencies
that remain
hidden

6


Dimensional model can show more
Extends definitions in
Logical and
Relational models –
additional 1:1 and 1:n
dependencies in the
scope of one table
can be described

7


Can be mapped to Logical or Relational
model
 Relational model – direct

mappings to tables and views
 Logical model – used entities
need to be engineered to
relational model in order to
have mapping to specific
tables

8


Dimensional model elements


Dimension



Level



Hierarchy



Link



Cube



Measure



Slice – subset of cube dimensions and measures

Oracle names (short, long, plural) can be defined for each element and exported to Oracle OLAP

9


Dimension


Can be Regular or Time



Consists of levels and hierarchies



Concept of merging dimensions is
supported – hierarchies can span
levels from more than one
dimension



Dimension has one base level that
defines dimension‟s grain

10


Merging dimensions example
 Level “Sales

Territory Region”
belongs to three
dimensions

11


Level
 Level is mapped to exactly one entity (table or view).

Several levels can be mapped to one entity
 Can be Regular level or level that represents value
based (parent-child) hierarchy
 Has a key defined over one or several attributes
 Has descriptive attributes that map to exactly one
column from dimension table

12


Level
 Mapping of optional short description and long description

attributes can be defined and mapping of mandatory „time
span” and “time end” attributes should be defined for level
belonging to time dimension
 Has calculated attributes – expressions over dimension
table columns
 Value based hierarchy is defined with “level key” and
“parent key” attributes
 Member selection criteria can be defined

13


Level - calculated attribute

14


Hierarchy
 Defines roll-up path and can connect levels belonging

to different dimensions (merging dimensions)
 Hierarchies are built and modified automatically during
the process of connecting levels and dimension node
 Three types of hierarchies – Parent-Child (value based
hierarchy), Level based and Ragged
 Joins are defined for levels mapped to different tables

15


Hierarchy
Hierarchy is created when dimension shape is connected (using “New Link”) to
base level.

16


Link
Used to define hierarchies and dimension usage. Child to parent attributes can be defined
when link connects levels defined on different entities

17


Value based hierarchy
Level is used to define key, parent key and descriptive attributes

18


Ragged hierarchy
Corresponds to skip level hierarchy in Oracle OLAP – use “New Ragged Hierarchy Link”
tool in order to mark and represent a “jump” over level(s) – dashed green link below

19


Dimension usage




20

link between cube and dimension – defines child to parent attributes (join of fact entity to base level
of dimension), whether it‟s role playing dimension and other related properties

Separate dimension is created in Oracle OLAP for each role playing dimension


Cube
 Defines fact entities and joins

between them
 Container for measures defined
on facts from fact entities
 Defines dimensions usage and
joins to dimensions – support for
role playing dimensions
 Container for “Slices” based on
cube definition

Cube defined on two fact entities:
FactInternetSales and
FactResellerSales

•
•
•

21


Abbreviations:
A – measure based on fact
FX – measure based on
formula expression
R – reference to dimension

Cube - partitioning
 Partitioning

dimension,
hierarchy, level
 Settings for

composites

22


Cube – summary levels (pre-compute)

23


Measure
 Measure can be based on single fact (direct mapping to

column) or expression involving facts and other measures
 Measure can be Full-Additive, Semi-Additive or NoneAdditive
 One (at least) or more aggregation functions can be defined
for Full-Additive and Semi-Additive measure.
 Semi-Additive measure has a list of dimensions on which it
is valid

24


Basic measure

25


Calculated measure

26


Reports
 There are no standard reports for Dimensional model
 Custom reports integrated in Search functionality can be

created – two levels of details are supported for cubes and
dimensions (in Data Modeler 4.0)
 Custom reports can be generated in HTML, PDF or Excel
format; the latter one can be used to change properties in
Excel file and then to import them back into model

27


Custom report template

28


Custom report – dimensions with details for
hierarchies and levels

29


Dimensional Model - usage
 Communication with customers – custom reports, detailed

and compact diagrams – for compact diagram use “Show
Compact model” in context menu for dimensional model
node in the browser
 Can be exported to Oracle AW
 Dimensions in Oracle physical model can be created using
“Engineer to Oracle Model” functionality

30


Detailed diagram

31


Compact diagram

32


Dimensions in Oracle physical model
 Used by database

engine for query
rewrite related to
materialized views
 Can be imported from
database dictionary or
DDL script

33


SQL Dimension

34


Dimension – physical model




35

Dimensions, levels and
hierarchies can be
created/edited
manually
Can be used as starting
point to create new
Dimensional model


Dimensional model – engineer from physical
model

36


Dimensional model – based on Sales History (SH)
sample schema

37


Example
Oracle SQL Developer Data Modeler 3.3 (4.0) is
distributed with sample design and complementary
description located in
datamodelerdatamodelersamplesDimensional_model
directory

38


Learn More..
 SQL Developer Data Modeler on OTN -

http://www.oracle.com/technetwork/developer-tools/datamodeler
 SQL Developer Data Modeler forum
https://forums.oracle.com/community/developer/english/development_to
ols/application_development_in_pl_sql/sql_developer_data_modeler
 Oracle Learning Library
http://apex.oracle.com/pls/apex/f?p=44785:2:0:FORCE_QUERY::2,RIR
,CIR:P2_TAGS:Data%20Modeler
 Jeff Smith http://www.thatjeffsmith.com/
 Ashley Chen http://oracletoolsgirl.blogspot.co.uk/
 Kris Rice http://krisrice.blogspot.co.uk/

39


40


41


Dimensional Modeling in Oracle SQL Developer Data Modeler

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (8)

Similar to Dimensional Modeling in Oracle SQL Developer Data Modeler

Similar to Dimensional Modeling in Oracle SQL Developer Data Modeler (20)

More from Jeff Smith

More from Jeff Smith (20)

Recently uploaded

Recently uploaded (20)

Dimensional Modeling in Oracle SQL Developer Data Modeler

Editor's Notes