A Practical Look at Managing Quality Within an EA...
Transcript of A Practical Look at Managing Quality Within an EA...
Managing Quality Within an EA
Repository
Version 01.00
Johan Schoeman
3
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Introduction
“Being good in business is the most
fascinating type of art” Andy Warhol, 1975
Philosophy of Andy Warhol (From A to B and Back Again)
RI Managing the Quality of an EA Repository v 01.0006/06/2008
4
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Introduction
“Architecture is the art of how to waste
space”Philip Johnson, 1964
New York Times, 27 December 1964
RI Managing the Quality of an EA Repository v 01.0006/06/2008
5
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Agenda
Case Study background
Challenges of repository quality management
Quality management process
Quality management reporting
Conclusion
RI Managing the Quality of an EA Repository v 01.0006/06/2008
6
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Case Study Background
Anglo Platinum is the world’s
largest primary producer of
platinum.
Anglo Platinum started with
their Enterprise Architecture in
2005 and has since invested a
significant amount in building up
content within the architecture
repository (ARIS), with
particular focus in the Business
Domain.
RI Managing the Quality of an EA Repository v 01.0006/06/2008
7
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Case Study Background (Cont.)
Real IRM provides outsource
EA services, including
repository management and
quality assurance.
Support to a number of
internal users, external
consultants distributed across
the Anglo Platinum operations.
RI Managing the Quality of an EA Repository v 01.0006/06/2008
8
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Agenda
Case Study background
Challenges of repository quality management
Quality management process
Quality management reporting
Ongoing challenges
RI Managing the Quality of an EA Repository v 01.0006/06/2008
9
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Controlling the content of the repository, becomes increasingly difficult
because of the amount of ARIS users and information in the repository.
Models and Objects numbers are growing significantly over time, causing complexity
in the amount of information to manage
Users are subject to turnover and get reassigned to working within new areas without
the knowledge of other users and administrators. When queries on content arise,
people struggle to find the person that is responsible for a particular area.
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Challenges for Repository Quality (Cont.)
Models 18207
Objects 99461
Users 80
10
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Challenges for Repository Quality (Cont.)
The variety of modeller backgrounds result in a variety of methodologies,
training and standards.
RI Managing the Quality of an EA Repository v 01.0006/06/2008
EmployeeRequired To
Attend To C...
EmployeeRequired To
Work Overtime
L4
Complete Pre-AuthorisedOvertime
Documentation
L4
Forward Pre-AuthorisedOvertime
Documentati...
L4
Determine IfPre-Authorised
OvertimeDocumentati...
L4 L3
Complete Pre-Authorised Overtime
Documentation
L3
Forward Pre-Authorised OvertimeDocumentation For
Capturing
L3
Determine If Pre-Authorised Overtime
DocumentationComplete
L3
11
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Challenges for Repository Quality
EA modelling tools provide limited control of input
RI Managing the Quality of an EA Repository v 01.0006/06/2008
SHE...
Rustenburg
Region
R
General...D
AnalyticalOccupatio...
Mine, Health
& Safety A...
OccupationalHygiene...
Area NoiseSurvey Report
SHE.14.03
L3
Analyse HygieneMeasurement
And MonitoringSamples
AreaIlluminati...
AnalyseHygiene...
Versi...
SIMRAC
Guideline
Mine
Manager...
A
Occupation
al Hygie...
E
Version 1.01
Mine Manager:Lower Mine...
A
ManagerEngine...
A
OccupationalHygiene...
OccupationalHygiene Iss...
OccupationalHygiene Ri...
OccupationalHygiene...
12
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Agenda
Case Study background
Challenges of repository quality management
Quality management process
Quality management reporting
Ongoing challenges
RI Managing the Quality of an EA Repository v 01.0006/06/2008
13
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process
RI Managing the Quality of an EA Repository v 01.0006/06/2008
In order to publish a model to the
production repository, the modeller
has to go through three stages:
Completing the EA model
Completing the Logical Quality
Assurance of the model
Completing the Technical Quality
Assurance of the model
Complete EAModel
Request LogicalQA of EA Model
Perform LogicalQA of EA Model
Logical QAPassed
Logical QA NotPassed
Yes No
RequestTechnicall QA of
EA Model
PerformTechnical QA of
EA Model
Technical QANot Passed
Technical QAPassed
NoYes
Publish Model inProductionRepository
CorrectTechnical QA
Errors
Simplified QA Process
14
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the EA Model Content
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Complete EAModel
Request LogicalQA of EA Model
Perform LogicalQA of EA Model
Logical QAPassed
Logical QA NotPassed
Yes No
RequestTechnicall QA of
EA Model
PerformTechnical QA of
EA Model
Technical QANot Passed
Technical QAPassed
NoYes
Publish Model inProductionRepository
CorrectTechnical QA
Errors
The modellers have various tools
available that help and guide them
in producing quality content:
EA Meta-model
EA Standards and Guidelines
EA Training
Capturing Interfaces
Assigning responsible modellers
15
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the EA Model Content
EA Meta-model
RI Managing the Quality of an EA Repository v 01.0006/06/2008
16
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the EA Model Content
EA Standards and Guidelines
RI Managing the Quality of an EA Repository v 01.0006/06/2008
17
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the EA Model Content
EA Training
To be completed by every modeller before they can work within the repository
Training not only teaches how to use the tool, but how to do modelling within Anglo
Platinum’s context, i.e. Meta-model, standards, guidelines, etc.
Training course includes a case study and evaluation, which needs to be passed
before repository access is allowed
RI Managing the Quality of an EA Repository v 01.0006/06/2008
18
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the EA Model Content
Capturing Interfaces
Modellers use a form-based
interface with which to capture
model content, which ensures that
the right type of information is
captured at the appropriate levels.
RI Managing the Quality of an EA Repository v 01.0006/06/2008
25
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the EA Model Content
Assigning responsible modellers
ARIS users are assigned as
responsible modellers for the models
within the repository
This helps other users and
administrators keep track of the people
responsible for models, when queries
arise.
This gets logged within a read only
attribute in ARIS to ensure that it does
not get changed erroneously
RI Managing the Quality of an EA Repository v 01.0006/06/2008
26
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the Logical QA
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Complete EAModel
Request LogicalQA of EA Model
Perform LogicalQA of EA Model
Logical QAPassed
Logical QA NotPassed
Yes No
RequestTechnicall QA of
EA Model
PerformTechnical QA of
EA Model
Technical QANot Passed
Technical QAPassed
NoYes
Publish Model inProductionRepository
CorrectTechnical QA
Errors
The Logical QA process ensures
that the model makes logical sense
and follows the logical guidelines. It
is essential for maintaining the
quality of the EA content.
Needs to be performed by a person
The Logical QA process consists out
of:
Requesting Logical QA
Performing Logical QA
Providing feedback on the Logical QA
27
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the Logical QA
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Requesting Logical QA
Modellers trigger a request for
Logical QA within ARIS
28
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the Logical QA
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Requesting Logical QA
Modellers trigger a request for
Logical QA within ARIS
This request logged into a read only
attribute within ARIS
29
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the Logical QA
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Performing Logical QA
Modellers trigger a request for
Logical QA within ARIS
This request logged into a read only
attribute within ARIS
This request is logged into an
external database that tracks the
QA history
QA DB
30
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the Logical QA
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Performing the Logical QA
The Quality Regulator evaluates the
model against a set of specified
criteria.
The result of the logical QA is
logged within a read only attribute in
ARIS.
The result is also logged into an
external database that tracks the
QA history.
QA DB
31
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the Logical QA
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Performing the Logical QA
Feedback is provided to the
responsible modeller, typicaly in
the form of a scanned model with
written comments.
This process continues until the
model passes the Logical QA
32
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the Technical QA
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Complete EAModel
Request LogicalQA of EA Model
Perform LogicalQA of EA Model
Logical QAPassed
Logical QA NotPassed
Yes No
RequestTechnicall QA of
EA Model
PerformTechnical QA of
EA Model
Technical QANot Passed
Technical QAPassed
NoYes
Publish Model inProductionRepository
CorrectTechnical QA
Errors
The Technical QA process ensures
that the model follows meta-model
rules and guidelines. This QA is
essential for repository integrity.
Is automated
The Technical QA process consists
out of:
Requesting Technical QA
Automatically performing Technical QA
Providing feedback on the Technical
QA
33
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the Technical QA
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Performing the Technical QA
The content of the ARIS repository
is exported to an external database.
Predefined SQL queries that check
for technical errors are run
ARIS DB
Export
DB
34
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: Completing the Technical QA
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Providing Feedback on the Technical QA
The result from the Technical QA is logged within ARIS and the QA DB.
Spreadsheets that highlight the technical errors are sent back to the applicable
modeller to correct.
35
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
QM Process: QA Completed
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Complete EAModel
Request LogicalQA of EA Model
Perform LogicalQA of EA Model
Logical QAPassed
Logical QA NotPassed
Yes No
RequestTechnicall QA of
EA Model
PerformTechnical QA of
EA Model
Technical QANot Passed
Technical QAPassed
NoYes
Publish Model inProductionRepository
CorrectTechnical QA
Errors
ARIS
Development DB
ARIS Production
DB
36
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Agenda
Case Study background
Challenges of repository quality management
Quality management process
Quality management reporting
Ongoing challenges
RI Managing the Quality of an EA Repository v 01.0006/06/2008
37
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Quality Management Reporting
RI Managing the Quality of an EA Repository v 01.0006/06/2008
QA DB
ARIS DB
Historical QA Data
Point-in-Time Data
REPORT A
REPORT B
REPORT C
REPORT D
Export
DB
38
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Quality Management Reporting
RI Managing the Quality of an EA Repository v 01.0006/06/2008
QA Status
Helps with tracking the Quality Assurance process over time
Row Labels <1 Month 1 Month
4-6
Months
6-12
Months
> 1
Year
Grand
Total
Logical Fail 1 5 144 112 262
Logical Pass, Technical Fail 48 62 3 98 3 214
QA Backlog 20 20
Logical QA in Progress 1 1
Logical QA Requested 11 11
Passed 23 136 159
Technical QA Requested 14 14
Grand Total 63 78 26 379 135 681
39
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Quality Management Reporting
RI Managing the Quality of an EA Repository v 01.0006/06/2008
QA Status
Provides QA statistics per modeller
Responsible Ave Logical Cycles
user1 1
user2 6
user3 2
user4 2
user5 1
Row Labels user1 user2 user3 user4 user5
Logical Fail 21 8 102 5
Logical Pass, Technical Fail 6 88 3 25
QA Backlog 14
Logical QA in Progress 1
Logical QA Requested
Passed 99 1
Technical QA Requested
Grand Total 27 210 105 30 1
40
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Quality Management Reporting
Modelling Progress
Provides management with a point in time view of the status of the EA modelling
projects.
RI Managing the Quality of an EA Repository v 01.0006/06/2008
41
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Quality Management Reporting
RI Managing the Quality of an EA Repository v 01.0006/06/2008
Modelling Progress
Provides a summarised overview of the growth of repository content.
42
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Agenda
Case Study background
Challenges of repository quality management
Quality management process
Quality management reporting
Conclusion
RI Managing the Quality of an EA Repository v 01.0006/06/2008
43
Copyright ©
Real IRM Solutions (Pty) Ltd
2008
Conclusion
Challenges are not individual to Anglo Platinum or ARIS alone, but are
very similar at a number of organisations and EA tools.
Using similar processes across clients with different toolsets
Quality Management is an ongoing and ever-increasing challenge and the
process and reporting has to mature over time
RI Managing the Quality of an EA Repository v 01.0006/06/2008