(otw13) agile data warehousing: introduction to data vault modeling
DESCRIPTION
This is the presentation I gave at OakTable World 2013 in San Francisco. #OTW13 was held at the Children's Creativity Museum next to the Moscone Convention Center and was in parallel with Oracle OpenWorld 2013. The session discussed our attempts to be more agile in designing enterprise data warehouses and how the Data Vault Data Modeling technique helps in that approach.TRANSCRIPT
Agile Data Warehouse Modeling:
Introduction to Data Vault Modeling
Kent Graziano
Data Warrior LLC
Twitter @KentGraziano
Agenda
Bio
What do we mean by Agile?
What is a Data Vault?
Where does it fit in an Oracle BI architecture
How to design a Data Vault model
Being “agile”
My Bio
Oracle ACE Director
Certified Data Vault Master and DV 2.0 Architect
Blogger: Oracle Data Warrior
Data Architecture and Data Warehouse Specialist ● 30+ years in IT
● 20+ years of Oracle-related work
● 15+ years of data warehousing experience
Co-Author of ● The Business of Data Vault Modeling
● The Data Model Resource Book (1st Edition)
Editor of “The” Data Vault Book
Past-President of ODTUG and Rocky Mountain Oracle User Group
Manifesto for Agile Software Development
“We are uncovering better ways of developing
software by doing it and helping others do it.
Through this work we have come to value:
Individuals and interactions over processes and
tools
Working software over comprehensive
documentation
Customer collaboration over contract negotiation
Responding to change over following a plan
That is, while there is value in the items on the right,
we value the items on the left more.”
http://agilemanifesto.org/
Applying the Agile Manifesto to DW
User Stories instead of
requirements documents
Time-boxed iterations ● Iteration has a standard length
● Choose one or more user stories to fit in that
iteration
Rework is part of the game ● There are no “missed requirements”... only
those that haven’t been delivered or
discovered yet.
Data Vault Definition
The Data Vault is a detail oriented, historical tracking
and uniquely linked set of normalized tables that
support one or more functional areas of business.
It is a hybrid approach encompassing the best of
breed between 3rd normal form (3NF) and star
schema. The design is flexible, scalable, consistent
and adaptable to the needs of the enterprise.
Dan Linstedt: Defining the Data Vault TDAN.com Article
Architected specifically to meet the needs
of today’s enterprise data warehouses
What is Data Vault Trying to Solve?
What are our other Enterprise
Data Warehouse options?
● Third-Normal Form (3NF): Complex
primary keys (PK’s) with cascading
snapshot dates
● Star Schema (Dimensional): Difficult to
reengineer fact tables for granularity
changes
Difficult to get it right the first
time
Not adaptable to rapid
business change
NOT AGILE!
(C) Kent Graziano
Data Vault Time Line
2000 1960 1970 1980 1990
E.F. Codd invented
relational modeling
Chris Date and
Hugh Darwen
Maintained and
Refined
Modeling
1976 Dr Peter Chen
Created E-R
Diagramming
Early 70’s Bill
Inmon Began
Discussing Data
Warehousing
Mid 60’s Dimension & Fact
Modeling presented by
General Mills and Dartmouth
University
Mid 70’s AC Nielsen
Popularized
Dimension & Fact Terms
Mid – Late 80’s Dr Kimball
Popularizes Star Schema
Mid 80’s Bill Inmon
Popularizes Data
Warehousing
Late 80’s – Barry
Devlin and Dr Kimball
Release “Business
Data Warehouse”
1990 – Dan Linstedt
Begins R&D on Data
Vault Modeling
2000 – Dan Linstedt
releases first 5
articles on Data Vault
Modeling
© LearnDataVault.com
Data Vault Evolution
The work on the Data Vault approach began in the
early 1990s, and completed around 1999.
Throughout 1999, 2000, and 2001, the Data Vault
design was tested, refined, and deployed into specific
customer sites.
In 2002, the industry thought leaders were asked to
review the architecture.
● This is when I attend my first DV seminar in Denver and met
Dan!
In 2003, Dan began teaching the modeling techniques
to the mass public.
(C) Kent Graziano
Where does a Data Vault Fit?
© LearnDataVault.com
Oracle Information Management Reference Architecture
Staging Layer ● Change tables
● Reject tables for Data Quality
● External tables for file feeds
Foundation Layer ● Transactional granularity
maintained
● Process neutral: no user or
business requirements
● Just recording what happened
Access and Performance
Layer ● Dimensional model
● “Star Schemas”
● Process specific: targeting user
and business requirements
Where does Data Vault fit?
Data Vault goes here
What is a Foundation Layer?
Basis for long term enterprise scale data
warehouse
Must be atomic level data
● A historical source of facts
Not based on any one data source or system
Single point of integration
Flexible
Extensible
Provides data to the access/reporting layer
(C) Kent Graziano
How to be Agile using DV and Oracle
Model iteratively ● Use Data Vault data modeling technique
● Create basic components, then add over time
Virtualize the Access Layer ● Don’t waste time building facts and dimensions up front
● ETL and testing takes too long
● “Project” objects using pattern-based DV model with OBIEE
BMM or Oracle Views
Users see real reports with real data
(C) Kent Graziano
Data Vault: 3 Simple Structures
© LearnDataVault.com
Data Vault Core Architecture
Hubs = Unique List of Business Keys
Links = Unique List of Relationships across
keys
Satellites = Descriptive Data
Satellites have one and only one parent table
Satellites cannot be “Parents” to other tables
Hubs cannot be child tables
© LearnDataVault.com
1. Hub = Business Keys
Hubs = Unique Lists of Business Keys Business Keys are used to TRACK and IDENTIFY key information
(C) Kent Graziano
Hub Definition
What Makes a Hub Key? ● A Hub is based on an identifiable business key.
● An identifiable business key is an attribute that is used in the source systems to locate data.
● The business key has a very low propensity to change, and usually is not editable on the source systems.
● The business key has the same semantic meaning, and the same granularity across the company, but not necessarily the same format.
Attributes and Ordering ● All attributes are mandatory.
● Sequence ID 1st, Busn. Key 2nd , Load Date 3rd ,Record Source Last (4th).
● All attributes in the Business Key form a UNIQUE Index.
© LearnDataVault.com
2: Links = Associations
Links = Transactions and Associations They are used to hook together multiple sets of information
(C) Kent Graziano
Link Definitions
What Makes a Link? ● A Link is based on identifiable business element
relationships. ● Otherwise known as a foreign key,
● AKA a business event or transaction between business keys,
● The relationship shouldn’t change over time ● It is established as a fact that occurred at a specific point in time
and will remain that way forever.
● The link table may also represent a hierarchy.
Attributes ● All attributes are mandatory
(C) LearnDataVault.com
Modeling Links - 1:1 or 1:M?
Today:
● Relationship is a 1:1 so why model a Link?
Tomorrow:
● The business rule can change to a 1:M.
● You discover new data later.
With a Link in the Data Vault:
● No need to change the EDW structure.
● Existing data is fine.
● New data is added.
(C) Kent Graziano
3. Satellites = Descriptors
Satellites provide context for the Hubs and the Links
(C) Kent Graziano
Satellite Definitions
What Makes a Satellite? ● A Satellite is based on an non-identifying business
elements.
● The Satellite data changes, sometimes rapidly, sometimes slowly.
● The Satellite is dependent on the Hub or Link key as a parent,
● Satellites are never dependent on more than one parent table.
● The Satellite is never a parent table to any other table (no snow flaking).
Attributes and Ordering ● All attributes are mandatory – EXCEPT END DATE.
● Parent ID 1st, Load Date 2nd, Load End Date 3rd,Record Source Last.
(C) LearnDataVault.com
Satellite Entity- Details
A Satellite has only 1 foreign key; it is dependent on the parent table (Hub or Link)
A Satellite may or may not have an “Item Numbering” attribute.
A Satellite’s Load Date represents the date the EDW saw the data (must be a delta set).
● This is not Effective Date from the Source!
A Satellite’s Record Source represents the actual source of the row (unit of work).
To avoid Outer Joins, you must ensure that every satellite has at least 1 entry for every Hub Key.
(C) LearnDataVault.com
Data Vault Model Flexibility (Agility)
Goes beyond standard 3NF • Hyper normalized
● Hubs and Links only hold keys and meta data
● Satellites split by rate of change and/or source
• Enables Agile data modeling
● Easy to add to model without having to change existing structures and load routines
• Relationships (links) can be dropped and created on-demand.
● No more reloading history because of a missed requirement
Based on natural business keys • Not system surrogate keys
• Allows for integrating data across functions and source systems more easily
● All data relationships are key driven.
© LearnDataVault.com
Data Vault Extensibility
Adding new components to
the EDW has NEAR ZERO
impact to:
• Existing Loading
Processes
• Existing Data Model
• Existing Reporting & BI
Functions
• Existing Source Systems
• Existing Star Schemas
and Data Marts
© LearnDataVault.com
Standardized modeling rules • Highly repeatable and learnable modeling technique
• Can standardize load routines
● Delta Driven process
● Re-startable, consistent loading patterns.
• Can standardize extract routines
● Rapid build of new or revised Data Marts
• Can be automated
‣ Can use a BI-meta layer to virtualize the reporting structures
‣ Example: OBIEE Business Model and Mapping tool
‣ Can put views on the DV structures as well
‣ Simulate ODS/3NF or Star Schemas
Data Vault Productivity
(C) Kent Graziano
• The Data Vault holds granular historical relationships.
• Holds all history for all time, allowing any source system feeds to be reconstructed on-demand
• Easy generation of Audit Trails for data lineage and compliance.
• Data Mining can discover new relationships between elements
• Patterns of change emerge from the historical pictures and linkages.
• The Data Vault can be accessed by power-users
© LearnDataVault.com
Data Vault Adaptability
Other Benefits of a Data Vault
Modeling it as a DV forces integration of the Business Keys upfront. • Good for organizational alignment.
An integrated data set with raw data extends it’s value beyond BI: • Source for data quality projects
• Source for master data
• Source for data mining
• Source for Data as a Service (DaaS) in an SOA (Service Oriented Architecture).
Upfront Hub integration simplifies the data integration routines required to load data marts.
• Helps divide the work a bit.
It is much easier to implement security on these granular pieces.
Granular, re-startable processes enable pin-point failure correction.
It is designed and optimized for real-time loading in its core architecture (without any tweaks or mods).
© LearnDataVault.com
Worlds Smallest Data Vault
The Data Vault doesn’t have to be “BIG”.
An Data Vault can be built incrementally.
Reverse engineering one component of the existing models is not uncommon.
Building one part of the Data Vault, then changing the marts to feed from that vault is a best practice.
The smallest Enterprise Data Warehouse consists of two tables: ● One Hub,
● One Satellite
Hub_Cust_Seq_ID
Hub_Cust_Num
Hub_Cust_Load_DTS
Hub_Cust_Rec_Src
Hub Customer
Hub_Cust_Seq_ID
Sat_Cust_Load_DTS
Sat_Cust_Load_End_DTS
Sat_Cust_Name
Sat_Cust_Rec_Src
Satellite Customer Name
© LearnDataVault.com
Notably…
In 2008 Bill Inmon stated that the “Data Vault
is the optimal approach for modeling the EDW
in the DW2.0 framework.” (DW2.0)
The number of Data Vault users in the US
surpassed 500 in 2010 and grows rapidly
(http://danlinstedt.com/about/dv-customers/)
Organizations using Data Vault
WebMD Health Services
Anthem Blue-Cross Blue Shield
MD Anderson Cancer Center
Denver Public Schools
Independent Purchasing Cooperative (IPC, Miami) • Owner of Subway
Kaplan
US Defense Department
Colorado Springs Utilities
State Court of Wyoming
Federal Express
US Dept. Of Agriculture
What’s New in DV2.0?
Modeling Structure Includes… ● NoSQL, and Non-Relational DB systems, Hybrid Systems
● Minor Structure Changes to support NoSQL
New ETL Implementation Standards ● For true real-time support
● For NoSQL support
New Architecture Standards ● To include support for NoSQL data management systems
New Methodology Components ● Including CMMI, Six Sigma, and TQM
● Including Project Planning, Tracking, and Oversight
● Agile Delivery Mechanisms
● Standards, and templates for Projects
© LearnDataVault.com
Conclusion?
Changing the direction of the river takes less
effort than stopping the flow of water © LearnDataVault.com
Summary
• Data Vault provides a data
modeling technique that
allows:
‣ Model Agility
‣ Enabling rapid changes and additions
‣ Productivity
‣ Enabling low complexity systems with high
value output at a rapid pace
‣ Easy projections of dimensional models
‣ So? Agile Data Warehousing?
Super Charge Your Data Warehouse
Available on Amazon.com
Soft Cover or Kindle Format
Now also available in PDF at
LearnDataVault.com
Hint: Kent is the Technical
Editor
Data Vault References
www.learndatavault.com
www.danlinstedt.com
On LinkedIn:
http://www.linkedin.com/groups?gid=44926
On YouTube:
www.youtube.com/LearnDataVault
On Facebook:
www.facebook.com/learndatavault
Contact Information
Kent Graziano
The Oracle Data Warrior
Data Warrior LLC
Visit my blog at
http://kentgraziano.com