Title: QI SymposiumExtending CMS Baseline with Data Warehouse
1(No Transcript)
2Part 1
- Data Warehouse for Newcomers
- (me!)
- Linda HoranCMS Data WarehouseInformation
Coordinator
3Data Warehouse Background-None
- December 2000
- Joined CMS Training/ Doc department
- March 2005
- Joined Data Warehouse group
- No Computer Science or Data Warehouse background
- Did the only thing I knew to do . . .
4Searched for Context . . .
- Armed with instructional technology toolkit, the
archaeological dig begins
5First Artifact - a Map
Strategic
Tactical
Operational
6A Working Definition - MW
- Data Warehousing 101
- A data warehouse is a large analytical database
which derives its data from a variety of
production systems and is structured for
querying, reporting, and analysis - by Arshad Khan
7Production (Operational) Systems
- Online Transaction Processing (OLTP)
- Flexible, fast transaction data capture
- Normalized databases
- Relational tables
8Normal, compared to what?
- Five Stages of Normal Forms
- Third Normal Form (3NF)
- No fields may depend on other non-key fields
- You dont need to know anything more unless you
are a data architect
9Its all Relative
- Method of structuring tables
- Associated to each other by shared attributes
- Tables are joined by Key(s)
- Primary
- Foreign
- Natural / Artificial
10Normalizing a Table
11Primary Key (PK) / Foreign Key (FK)
12Artificial Key / Surrogate Key
13Transaction System Benefits
- OLTP good for
- Rapid single record update
- Rapid single record retrieval
- Little risk of introducing inconsistencies
- Whats good for OLTP is not good for reporting
14Job Report Example
15Report with JOB and EMPLOYEE Data
16Reporting from Production Systems
- Most reports require data from many tables
- Ad-hoc reports are difficult to create
- Programming tools required
- Data is stored in multiple applications
- Performance degrades!
17Reporting / Analytical Systems
- Reporting needs are opposite of OLTP
- Denormalized
- Optimized for reporting and analytics
- Speeds data retrieval
- Not intended for data entryand modification
18Denormalized
- Redundancy purposely introduced in favor of query
performance
19Extract, Transform, Load (ETL)
20Dimensional Model
- The Dimensional model is based on a structure
organized by dimensions - Fact Table
- Contains measurements or metrics of facts of
business processes - Dimension Table
- Who, what, where, when, and how of a measurement
(Fact)
21Synonym Star Schema
22Star Schema
- Fact Student Career Term
- Dim Institution
- Dim Term
- Dim Academic Program
- Dim Person
- And others . . .
23Conformed Dimensions
- A dimension that has exactly the same meaning and
content when it is referred from different fact
tables - PS_D_SA_PERSON
- Admissions
- Student Financials
- Etc.
24Metadata
- Data about data
- Structural metadata refers to information about
table structure - Other uses of metadata
- Data definitions
- Transformations
- Date of last update
25Online Analytic Processing (OLAP)
- OLAP Processing that supports the analysis of
business trends and projections - OLAP CUBE Dataset organized in a hierarchical,
multidimensional arrangement
26Data Mining - Top of the Pyramid
- Data mining finds patterns and subtle
relationships in data and infers rules that allow
the prediction of future results. - Generally implemented using OLAP cubes
27Other Terms of Confusion . . .
Operational Data Store
Data Mart
BUSINESS INTELLIGENCE
HistoricalReporting
Slowly Changing Dimensions(all three types!)
28X Marks the Spot
This is all BI!
Exec
VP Level
Operational Mgt.
X
Line Mgt HR Mgt Benefit Mgt
Daily Transactions
HR Recruitment Benefits Payroll TL
29Its Show Time!
- "In theory there is not much difference between
theory and practice. In practice, there is." - Yogi Berra
30Part 2
Introduction
- The CMS Data Warehouse Project
- Donna Furon
- CMS Data WarehouseProject Director
31Goals
Introduction
- Why is there a CMS Warehouse?
- What is provided?
- How is it working?
- What is next?
- Questions and Answers
32Goals
Goals
- Why is there a CMS Warehouse?
- What is provided?
- How is it working?
- What is next?
- Questions and Answers
33History
Why is there a data warehouse?
- Data Warehouse pulled from CMS scope
- Campus Efforts (SLO, Northridge, RDS)
- Data Warehouse Consortium
- CMS EC calls the question
34Need CMS Project Directors
Why is there a data warehouse?
- Immediate
- Operational reporting
- Integrated reporting
- Long Term
- Historical reporting
- Analytics
35Need Operational Reporting
Why is there a data warehouse?
- Transaction system complexity
- Many tables and lookups
- Cryptic names
- Cryptic codes
- CPU contention
- Tuning methodologies
36History Approach 6/2003 11/2003
Why is there a data warehouse?
- CMS Central
- License Software Tools and Train Campuses
- Develop, Support and Maintain Baseline
- Campuses
- Acquire, Support and Maintain Hardware
- Incorporate and Support Local Modifications
37Approach Assumptions
Why is there a data warehouse?
- Economies of scale
- Purchasing software
- Developing baseline
- Campus needs
- Hardware flexibility
- Customization ability
38History Feasibility Study 11/2003 5/2004
Why is there a data warehouse?
- Recommended Approach versus Individual Campus
Efforts - 101 Savings
- Sunk Costs Considered
- EC Authorized RFP
39Goals
Goals
- Why is there a CMS Warehouse?
- What is provided?
- How is it working?
- What is next?
- Questions and Answers
40History RFP
What is provided?
- Complete Solution
- Tools
- Data Models for PeopleSoft Source
- Include all three PS Apps
- Train CMS Central to Distribute
- Bundle as a Single Solution
- Multi-vendor bids OK
41History Oracle/PeopleSoft Bid
What is provided?
- EPM with JumpStarts
- Extended IBM (formerly Ascential) DataStage
Licensing - Authorization to Distribute Code Outside of EPM
- Multiple Platforms
- Windows/Oracle Added
42The Data Warehouse Team
What is provided?
- Project Director
- Information Coordinator
- Lead Developer
- Two Developers
- SOSS application and HOSS technical team support
- Campuses and Chancellors Office
43Consultant Deliverables
What is provided?
- Oracle (PeopleSoft) vanilla warehouse
- Content for all modules in
- Human Resources 8.0
- Student Administration 8.0
- Finance 8.4
44Consultant Deliverables
What is provided?
- CMS DW team training/mentoring
- Worked side-by-side
- A lot of questions
- Modeled
- HR Time and Labor
- Finance FNAT, AAT, and Fund
45Campus Receives
What is provided?
- Tools, Training, and Technology
- IBM DataStage Training
- Run and monitor jobs
- Install baseline updates/upgrades
- Customize
- Delivered project installed by DW team
- Report writer overview/workshop
46Deliverables - Implementation
What is provided?
- Kickoff meeting
- Environment validation
- Installation by DW Team
- Jobs to create warehouse tables
- Jobs to refresh warehouse tables
47Deliverables - Implementation
What is provided?
- Training IBM DataStage
- Run and monitor jobs
- Install baseline updates/upgrades
- Customize
48Deliverables - Implementation
What is provided?
- Workshop / orientation for report writers
- Documentation
- Data warehouse specific
- Star schema design
- Surrogate IDs (SIDs)
- Dimension and Fact Tables
- Current dimensions
49Deliverables - Implementation
What is provided?
- Workshop / orientation for report writers
- PeopleSoft specific (e.g. EFFDTs)
- Create sample report (class exercise)
- Workshop assumes
- Campus application knowledge
- Campus BI tool knowledge
-
50Deliverables Ongoing Support
What is provided?
- Updates for application updates
- Projects and code for new fields
- Upgrades
51Campus Receives
What is provided?
- Timely updates
- Evaluating patches and fixes
- Upgrades
- Access to repositories
- Help Desk support
52Deliverables Repositories
What is provided?
53Campus Responsibilities
Campus Responsibilities
- Hardware support
- Customization / change control
- Updates and upgrades from baseline
- Reporting
54Goals
Goals
- Why is there a CMS Warehouse?
- What is provided?
- How is it working?
- What is next?
- Questions and Answers
55Pilot Campus Implementations
How is it working?
- San Luis Obispo (HR) 11/2005
- Chancellors Office (Fin) 12/2005
- Chico (Fin) 1/2006
- Pomona (SA) 2/2006
- Sonoma (SA) 2/2006
56Lessons Learned
How is it working?
- Process validated
- Isolate training
- Improve performance
- Add CSU functionality
57Goals
Goals
- Why is there a CMS Warehouse?
- What is provided?
- How is it working?
- What is next?
- Questions and answers
58Scope Near Term
What is next?
- Add CSU Functionality
- 8.9 Upgrades
59Scope Long Term
What is next?
- Historical Reporting
- Analytics
- Fusion
60CSU Functionality
What is next?
- Data Warehouse Advisory Group
- Representatives User Groups, Systemwide, IRDs,
HOSS, SOSS - Prioritize CSU functionality additions
- Recommend future functionality
61Training
What is next?
- Sessions scheduled at the CO
- Training documentation updated
- Web-FAQs in progress
62Performance Tuning
What is next?
- Activities underway by staff
- IBM technical support calls
- Use of hash files
- IBM Consulting
- Look at code
- Recommendations software and architecture
63Campus Implementations
What is next?
- CMS Data Warehouse is core 10/1/2008
- Training week at the CO
- Campuses being scheduled now
- One week campus implementation 1st
- 2nd and 3rd implementations shorter?
648.9 Upgrades
What is next?
- Human Resources 3/2006 6/2006
- Student Administration to begin no later than
7/2006 - Finance 9/2006 12/2006
- Consulting being used
65Summary
Summary
- CMS Data Warehouse provides
- Systemwide, supported baseline
- For operational reporting
- With campus flexibility
66Summary
Summary
- CMS Data Warehouse is
- A supported basis for campus operational
reporting - An opportunity to leverage knowledge across
campuses - In a cost-effective model
67Goals
Questions and Answers
- Why is there a CMS warehouse?
- What is provided?
- How is it working?
- What is next?
- Questions and Answers
68For more information
Information
- CMS Webhttp//cms.calstate.edu
- Applications / CMS Data Warehouse
- Donna Furon dfuron_at_calstate.edu562.951.4319
- Linda Horan lhoran_at_calstate.edu562.951.4241
69(No Transcript)