You are on page 1of 29

ArchitectingADataWarehouse:

ArchitectingADataWarehouse:
A CaseStudy
ACaseStudy
A
Case Study
Project:zBis
CarlZeissVisionNorthAmerica

MarkGinnebaugh,UserGroupLeader,
Mark
Ginnebaugh User Group Leader
mark@designmind.com

TheJourney
DeterminedNeedforEnterpriseDataWarehouse
Determined Need for Enterprise Data Warehouse
WorkedwithBusinessUserstoUnderstandBusiness

Requirements
DeterminedSoftwareRequirements
D
i dS f
R
i
MSSQLServer2005&2008
MSSSIS(ETLTool)
MS SSIS (ETL Tool)
MSSSAS(AnalyticCubeTool)
MSSSRS&Excel(ReportingTools)
SharePointforDeployingReportsoverCompany
Intranet
DesignedandDevelopedzBis
g
p
DataWarehouse

ZBIS=WhatWeWillDeliver

Thisprojectteamwilldeliverthefollowing:
ConsolidatedreportingforCarlZeiss Vision
X

NorthAmerica

X
Reportingthatisconsistentandfromone
datawarehouse

X
Reportingthatiseasytouseandeasyto
R
ti th ti
t d
t
access

X
Toolsetwillbeflexibleandabletogrowand
changewithourbusiness

X PhaseIrocksoliddownloadfromERP/Manf

g abilitytoreviewlabinformation
y
Providing
asalabnetwork notindividualsilos with
accuratereportingacrossallproductsand
services
Wewilldeliverthebestproductpossiblebasedontheinformationwe
canplaceinourdatawarehouse!

Reporting from cubes off source systems only


No
N data
d t warehouse
h
Disparate
p
data systems
y
with different results from
each
Most systems not balanced to GL
Reporting for each business unit only
No reporting across all business units

TransactionalCubeofApproach
SalesQueries

OtherReports

SalesReports

Corporate

Download
D
l d
DataMart
Finance

DataMart
Inventory

DataMart
Sales&Marketing

ETLLoads
ETLLoad

ODS/Staging
g g
OperationalDataStore
ETLLoad

ERP

Manufacturing

Other

BI Tools/Analytics

Active
Reports

Excel

Static Reports

PerformancePoint Server
SharePoint
SQL
Reporting
Server

SQL Analytics
Server (SSAS)

Aggregated
Data Mart
TBD

Finance
Data Mart

Inventory
Data Mart

ETL Load (SSIS)


Data Warehouse

ETL Load (SSIS)


O S/S
ODS/Staging
Operational Data Store

ETL Load (SSIS)


ERP

Manufacturing SW

Other Data Sources

Sales
Data Mart

IntroductiontoDataWarehousing
WhatisaDataWarehouseSystem
WhyaDataWarehouseVs.CubesonSourceSystems
y
y
StarSchemaVs.TransactionalDataWarehouses
StarSchemaseaseofsystemintegrating
StarSchemasprovidesubstantialperformancegains
StarSchemashierarchycapabilitiesorDrillDown

Capabilities
Capabilities
RalphKimballDevelopedCurrentIndustryStandardsforStar
Schema DimensionsandFacts

DataWarehouseProjectLifecycle
Technical
Architecture
Design
Project
Planning

Business
Requirement
Definition

Dimensional
Modeling

Product
Selection &
Installation
Physical
Design

Report
Specifications

Data Staging
Design &
Development
Report
Development

Testing
ETL &
DW/DM
Report
Testing

Project Management

Deployment

Maintenance

4 + 1 Steps
Steps
4+1
DimensionalDesignProcess
RalphKimballsProcessforDevelopingStarSchemas
1. DetermineBusinessProcess

2
2.

Identify the Grain of the Fact


IdentifytheGrainoftheFact

3.
4.

ModelbusinessProcesses
Model
business Processes
EachProcesswilldetermine1ormoreFacts
DesignDWbyBusinessProcessNotBusinessUnit
Whatdoes1rowinFacttablerepresent
TransactionalorSummary

DesigntheDWDimensions
D
i
h DW Di
i
DesigntheDWFacts

+1 DetermineHierarchies
Determine Hierarchies

BusinessDrivenvs.DataDriven
DesignDW/BISystemviaBusinessProcess
DevelopDW/BISystemviaDatafromSourceSystems
l
/
f
ProfileDataasearlyaspossible
UnderstanddataanddesignDWusingexistingdata
Understand data and design DW using existing data

Design&DevelopusingbothBusinessProcessandavailable
Design & Develop using both Business Process and available

Dataifpossible

UnderstandingYourBusiness
IdentifykeybusinesssponsorsforDWproject
UseCorporateOrgChart

Setupinitialinterviewswithkeysponsors
DevelopBusinessProcessdiagrams
DevelophighlevelUseCaseDiagrams
D l hi h l l U C
Di
DetermineBusinessHierarchies

The Business Executive Interview


TheBusinessExecutiveInterview
Whataretheobjectivesofyourorganization?
WhatBusinessgoalsdoyouwanttoaccomplishwiththe

d l
developmentofzBis
t f Bi datawarehouseSystem?
d t
h
S t ?
Howdoyoumeasuresuccess?Howdoyouknowyouaredoing
How do you measure success? How do you know you are doing

well?Howoftendoyoumeasureyourcorporateperformance?
Whatareyourkeybusinessissuesthatyouaretryingtosolve

fromthezBis system?Iftheseissuesarenotjustifiedwhatisthe
impact to your department and organization?
impacttoyourdepartmentandorganization?

The Business Executive Interview


TheBusinessExecutiveInterview
Howdoyouidentifyproblemsorknowwhenyoumightbe

headedfortrouble?

Howdoyouspotexceptionsinyourbusiness?What

opportunitiesexisttodramaticallyimpactyourbusinessbased
opportunities
exist to dramatically impact your business based
onimprovedaccesstoinformation?Whatisthefinancial
impact

Ifyoucould..,Whatwoulditmeantoyourbusiness?
Whatisyourvisiontobetterleverageinformationwithinyour
What is your vision to better leverage information within your

organization?

Howdoyouanticipatethatyourstaffwillinteractdirectlywith
H d
ti i t th t
t ff ill i t
t di tl ith

thisinformation?

Th B i
M
I t i
TheBusinessManagerInterview
Whataretheobjectivesofyourdepartment?
What are the objectives of your department?
Whatareyoutryingtoaccomplish?Howwoulddoyougo

about achieving your objectives?


aboutachievingyourobjectives?
Whatareyoursuccessmetrics?
Howdoyouknowyouaredoingwell?
Howoftendoyoumeasureyourdepartment/team?
y
y
p
Howdoyouanticipatethatyourstaffwillinteractdirectlywith

thisinformation?

BusinessProcessDiagrams
UnderstandBusinessRequirementsforbuilding

DW/BI system.
DW/BIsystem.
DefinestheMeasuresandDimensionsfordata
Defines the Measures and Dimensions for data

warehouse

DetermineHierarchies
CustomerHierarchies SalesChannels
DistributionChannels
BusinessChannels
CustomerChannels
ProductDivisions
Product Divisions
SalesOrganizations
SalesOffice
Sales Office
BuyGroups/DirectlyPurchase

DetermineHierarchies
ProductHierarchy
Manufacturer
Brand
ProductType EachproducttypehadownHierarchy

Lens
Service
Equipment
etc
t
Design
Make/Model
/

DetermineHierarchies
GeoHierarchy
SalesDivision
SalesRegion
SalesTerritory

ConformedDimensions
Standardizeddimensionsacrossdatawarehouse
St d di d di
i
d t
h
Dimensionsareassociatedwithmultiplebusiness

processes
DeterminebyusingBusMatrix&enforcedinETL
ConformedDimensionsaresharedandconsistent
C f
d Di
i
h d d
it t
acrossfacttables

UseDataWarehouseBUSMatrix
UseDataWarehouseBUSMatrixfor
Understanding&mappingofBusinessProcessesand

Dimensions
OngoingDW/BIplanningefforts
Team&ManagementCommunications
Team & Management Communications
UnderstandBusinessProcessunionsacrosstheenterprise

DataWarehouseBUSMatrix
Date

Company

Customer

Product

Geo

DistCtr

Promo

Company
Sales

Customer
Discounts

Product
Cost

Company
Inventory

DistCtr
Inventory

De elop Dimensional Schema


DevelopDimensionalSchema

Sl Ch i Di
i
SlowChangingDimensions
Type1
Type 1 OverwriteexistingDimensionRow
Overwrite existing Dimension Row
Usewhendontneedtokeephistorydatarow
Canbeusedtocorrectbaddata

Type2 CreateanewDimensionRow
Usedateand/oractivenonactivefieldstoidentifycurrent

andinactivedatarows
Type3 KeepoldandaddnewattributesinDimensionRow
AllowAlternaterealitiestoexistsimultaneouslyinone
Allow Alternate realities to exist simultaneously in one
DimensionRow
SlowChangingDimensionsarehandledintheETL

T
f Di
i
TypeofDimensions
MiniDimension
Mini Dimension
JunkDimensions
OutriggerDimensions
Outrigger Dimensions
SmallStaticDimensions
Lookuptables
Lookup tables

T
fF t
TypeofFacts
TransactionFactTables
SnapshotFactTables
AccumulatingSnapshotFactTables
ConsolidatedorAggregatedFactTables

B id T bl
BridgeTables

B id T bl
BridgeTables

R
d d R di li
RecommendedReadinglist
TheDataWarehouseToolkit:TheCompleteGuidetoDimensional

Modeling(SecondEdition)
Ross
M d li (S
d Edi i ) byRalphKimballandMargy
b R l h Ki b ll d M
R
TheMicrosoftData WarehouseToolkit:WithSQLServer2005andthe

MicrosoftBusiness IntelligenceToolset byJoyMundy,Warren


Thornthwaite,andRalphKimball
BuildingaDataWarehouse:WithExamplesinSQLServer(Expert
Building a Data Warehouse: With Examples in SQL Server (Expert'ssVoice)
Voice)

byVincentRainardi
TheDataWarehouseLifecycleToolkit byRalphKimball,Margy Ross,

WarrenThornthwaite,andJoyMundy
TheDataWarehouseETLToolkit:PracticalTechniquesforExtracting,

Cleanin byRalphKimballandJoeCaserta
by Ralph Kimball and Joe Caserta

Tolearnmoreorinquireaboutspeakingopportunities,
pleasecontact:
MarkGinnebaugh,UserGroupLeader
mark@designmind.com

You might also like