SQL Server Analysis Services Understanding Unified Dimension Model (UDM)

Post on 09-Feb-2016

30 views 0 download

Tags:

description

SQL Server Analysis Services Understanding Unified Dimension Model (UDM). Agenda. Introduction to UDM Dimension model core components Dimension model example Dimension tables Star and Snowflake schemas Cubes Cube storage UDM Components UDM Overview. Introduction to UDM. - PowerPoint PPT Presentation

Transcript of SQL Server Analysis Services Understanding Unified Dimension Model (UDM)

SQL Server Analysis Services

Understanding Unified Dimension Model (UDM)

Agenda

»Introduction to UDM

»Dimension model core components

»Dimension model example

»Dimension tables

»Star and Snowflake schemas

»Cubes

»Cube storage

»UDM Components

»UDM Overview

Introduction to UDM

»The SSAS mission is to break out of the OLAP space by unifying the relational and dimensional reporting models using UDM.

Dimension model core components

»Measures - represent the numerical values (facts) that are used to measure business activity.

»Dimensions - the main goal of the dimensional model is to allow users to slice and dice data using different perspectives called dimensions.

»Dimension hierarchies - to facilitate drilling through data, dimensions may have hierarchies.

»Dimension members - the actual dimension entities that belong to each level are called dimension members.

Dimension model example

Dimension tables

»Dimension data are stored in relational tables called dimension tables.

‣usually wide (have many columns) but don’t have many rows

‣large number of columns is required to accommodate various dimension-related attributes

»The classic dimensional model defines two types of relational schemas that describe the relationship between the dimension and fact tables:

‣Star schemas

‣Snowflake schemas

Star and Snowflake schemas

»A star schema requires that a dimension hierarchy be contained within a single table. This requires the dimensionaltable to be denormalized.

»If the dimension hierarchy is left normalized, then the schema is of a snowflake type schema.

Cubes

»The Cube is logical storage object in SSAS.

»Cube combines dimensions and measurements to provide fast multidimensional access to cube data.

Cube storage

»Stores all fact rows and have aggregated values for each cell

»Cube aggregation options

‣ROLAP - the fact data and aggregations will be kept in the relational database.

‣MOLAP - the fact data and aggregations will be kept in the cube

‣HOLAP - the fact data and aggregations will be kept in the relational database and cube.

»SSAS usually uses MOLAP cubes.

»MOLAP storage option requires cube processing.

UDM Components

»Data source view (DSV) - logical data schema that seeks to present the data from the relational datastore in a standard and intuitive way.

»Dimensional model – cubes

»Calculations - business logic.

»End-user model - provides intuitive end-user reporting and data navigation experience. (reports, KPI)

»Management settings - cube configurations to meet various operational requirements, including availability, latency, and security

UDM Components

UDM Overview

»UDM goal to unite the best of both worlds (relational and dimensional) and become a bridge between the users and data.