Вы находитесь на странице: 1из 27

Data Warehousing for the

Future @ CQU
Ken Diefenbach
Acting Associate Director
Corporate Information Systems
Information Technology Division

.
05/29/1622/07/2004
05/29/1622/07/2004

Presentation Sequence
Some Thanks
Setting the Scene
CQU itself
BIG Guys

The Need Perceived


The Journey to the Warehouse Begins
Speedbumps Unlimited

The Warehouse Journey


Current State of Play
Future Directions
Questions / Comments
05/29/16

Setting the Scene


Central Queensland University
Geographically diverse and
getting diverser
16K EFTSL, 21K students
1k+ staff
2004 Budget ~ 120M
Major corporate systems:
PeopleSoft Financials & Higher
Education
Concept HRMS

Usual plethora of feral systems

05/29/16

Setting the Scene


Business Intelligence Group

(Think BIG)
3 members, 80+ years IT experience
Impromptu / PowerPlay / IWR
Site licenses for Impromptu & PowerPlay client software
1500+ web searchable reports

Moi

15 years + COGNOS tools


VCs Excel award 2001 (largely for Impromptu access to
PeopleSoft Student data)
Once again Member of International HEUG Reporting & Analysis
Product and Advisory Group

Prone to talk too much and too fast

Please tell me if its happening again!

05/29/16

What is a Data Warehouse?


NOT a copy of the transaction data base
on a separate server
A repository where information from
multiple sources is restructured and stored
in a format more suitable for analysis and
reporting
CQU is adopting the Warehouse Gospel a
la Ralph Kimball
05/29/16

A Useful Analogy
Reporting from
Production is like
counting screws on a
rapidly moving assembly
line
The ODS is like counting
screws on a stopped
assembly line
In a warehouse, we count
the screws in a single box
called screws
05/29/16

The Need Perceived


Impromptu catalog and reports very successful
No cross application reports, some reports impossible to
write
Use of dynamic / smart hot files widespread to overcome
the length of navigation required
Created an proof of concept ODS with assistance from
Oracle DBA guru turned academic (beware the dark
side)
Copy of subset of tables, views materialised for faster reporting

Reports run against either instance, but some reports


(esp with outer join requirements) still difficult /
impossible
05/29/16

The Need Perceived


Information stovepiped
No big picture / KPIs / Performance
Monitoring / Scorecarding
We had pushed Impromptu (our reporting
tool) as far as it or any other product
available would go against our Corporate
Systems, and there were still gaps which
needed addressing
05/29/16

Our Sales Pitch to Users


Faster reporting from the data
warehouse
Faster access to the originating
applications
Better quality data
Value added data providing
information that does not exist in
the source applications
Ability to report across
applications
05/29/16

Benefits ( more)
Provide a single version of the truth
Have the ability to import additional data
Provide better data structures for ad hoc
reporting
We can do PowerPlay (OLAP) stuff again!!

05/29/16

10

Limitations
Not necessarily the latest view of the data
Development will be evolutionary, not big
bang
Data must already exist or be derivable

05/29/16

11

The Journey Begins


Data free zone
(Acting Pro Vice Chancellor circa 2002)

Lonely Vice-Chancellors Lunch great opportunity to


start the push
(Student-free zone New Year 2004)

Presentations to Senior Management about the needs


assistance from COGNOS
Build the foundation first

Minimise risk
No big ticket expenditure (existing resistance)
Not interested in Big Bang approach (existing resistance)
Much prefer slowly slowly catchee monkey (survival strategy)

05/29/16

12

Speedbumps Unlimited
Business case from hell
Executive turnover
Budget blues
Politics
Business Case --> RFI --> ITO
Delays due to (un)due process and staff
unavailability dragged on well into 2004
Extended evaluation / selection process
05/29/16

13

At last the Decision!!


4 products shortlisted after proof of
concept & demonstrations
3 would do the job from our observations
2 within budget
DecisionStream was the only one that did
both

05/29/16

14

Why DecisionStream?
Dimensionally friendly
Dimensionally friendly
Slowly Changing Dimension Friendly
Strong vendor relationship (over 20 years)
Headstart for PS Financials
Starter (err throwaway) Impromptu & PowerPlay
Catalogs and Models
#$^&*()(*&^%$ cheap err good value
And did I already mention it handles dimensional
processing?
05/29/16

15

The Journey Begins in Earnest


Product purchased in late May 04
Training and consulting arranged for July 04
First requirements gathering run to
Melbourne / Sydney / Brisbane / Gold Coast
campuses in August 04
Continual dialogue with power users

05/29/16

16

The Journey Continues


Requirements gathering also a
promotional activity
Also expectation setting exercise
Promised industrial strength ODS in 6
weeks
Promised first release of data marts out by
end of year
Actually Mid June 2005
05/29/16

17

Complications
Everyone wants to dump their old data in
the warehouse
DecisionStream has more switches than a
electrical wholesale shop and more
options than a Lego catalogue
Executive turnover + University review +
Reorganisation + ..
Meant ever moving interview target
05/29/16

18

More complications
Personal Issues 2004

Issues beyond our / my control

Deployment pushed out BUT

Additional time spent learning what the tool does and does not
do we believe will continue to pay dividends

Personnel Issues 2005

Shift happens
Acting Associate Directorship limited Powerplay expertise
availability
When modelling begins in earnest, expect a bit of extra work on
the ETL end

Team at full strength numerically End May 2005 but need


to bring new member up to speed
Sometimes staff called in to help in other areas wed
prefer to avoid (eg debugging of complex PS COBOL)
05/29/16
19

From my Perspective
Conscious decision to focus first on investigating
technology and understanding warehousing
Revisited schedule after Christmas break 2004-5 and
redeployment
We all know what we cant answer

Lots of street credibility in the bank


Beta release of first two marts out with power users
very pleased with the results

First query on GPA at Melbourne Campus

timed out after 30 minutes on production / and on ODS


21 seconds on first mart at a finer grain level (student course
enrolments)
5 seconds on aggregated mart (student current program level)

Currently takes around 1 second!!!

05/29/16

20

From the Unis Perspective


Users reluctant to move off prod system
Technical issues with MOE and individual
TNSNAMES file

Will be encouraged to do so
Carrot (speed of ODS) and
Stick (cut down time out length on production system)

As warehouse marts delivered, the resulting


speed improvement will be impossible to ignore

05/29/16

21

Where next?
Requirements gathering & consolidation
Deploy more student data first
Move to other and the cross-subject areas
(Finance & HRMS in particular)
ReportNet / Scorecarding / Dashboarding /
Case Management ???

05/29/16

22

Diving in for a Closer Look

05/29/16

23

Stars in Our Eyes


We chose Kimball Star
Schema approach as it
was our preferred of 2
approaches available
cf Bill Inmons
approach
Dimensional Tables
shared across all Fact
Tables (Conformed)
05/29/16

24

Student Program Data Mart For


Every Student Program

Total Courses Registered


Total Courses Dropped
Total Courses Enrolled
Total Courses Passed
Total Courses Failed
Total Courses Outstanding

Total Credit Points Attempted


Total Credit Points Passed
Total Credit Points
Outstanding

05/29/16

Weighted GPA
Total Grade Points
Total GP Credits Taken
Total Workload Attempted
Total Workload Passed
(EFTSL)
Total Workload Outstanding
(EFTSL)
CQU Transfer Credits
Other Transfer Credits
Total Transfer Credits

25

Student Program Dimensions


Student Personal Details

Personal information about each student, including names and addresses,


biodemographic.data, citizenship and presence at the University.

Program Details

Information about each academic program.

Latest Student Program Details

The latest information about each student program.

Expected Final Term Details

The term the student expects to finish the program.

Completion Term Details

The actual term the student completed the program.

First Student Program Details

The earliest information about each student program.

Application Details

Some information about the student program application.

05/29/16

26

Questions?

05/29/16

27

Вам также может понравиться