FUJ00001377 - ICL Pathway Generalised API for OPS/TMS v1 Appendix C System Management

Evidence on official site

ICL Pathway

FUJ00001377

FUJ00001377
Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

Document Title:

Document Type:

Release:

Abstract:

Document Status:

Author & Dept:

Contributors:

Reviewed By:

Comments By:
Comments To:
Distribution:

Generalised API for OPS/TMS: Appendix C System
Management

Technical Design Standard

N/A

This appendix describes System Management interfaces to
be used, and functions that must be supported by
applications developed for the Horizon environment.

The main document provides the information required to plan
the development of new applications and describes in more
detail the architecture set out in the OPS Architecture
Specification.

Both documents are supplied under the terms of the Codified
Agreement to POCL to facilitate the procurement of
applications to run on the Service Infrastructure (interfacing
with OPS and TMS).

This document is only available to organisations outside ICL
Pathway through formal Non-Disclosure Agreement.

APPROVED

Jon Cruise, Technical Design Authority

Tony Hayward, Janet Dore

ICL Pathway: Terry Austin, John Dicks, John Allen, Simon
Fawkes, Allan Hodgkinson, Glenn Stephens, James
Stinchcombe, Dave Tanner, Peter Wiles

POCL: Bob Booth

N/A

Document Controller & Authors

ICL Pathway Library and Reviewers

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-1 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
0 Document control
0.1 Document history
Version [Date Reason for Issue (Associated
No. ICP/PinICL No.
0.10 First version of this appendix.
0.11 7/2/2000 _ISecond version
0.12 24/2/2000 IComments from earlier versions incorporated.
0.13 29/2/2000 IComments from earlier versions incorporated.
0.14 21/3/2000 IComments from earlier versions incorporated.
0.15 [30/3/2000 IComments from earlier versions incorporated
0.16 (3/4/2000 IComments from earlier versions incorporated.
0.17 17/4/2000 IComments from earlier versions incorporated.
0.18 3//5/2000 IComments from earlier version incorporated.
1.0 5/5/2000 _ [Issued for approval.

0.2 Approval authorities

Name Position Signature [Date
T. Austin Development Director

J. Dicks (Customer Requirements Director

R. Booth POCL

© 2000 ICL Pathway Limited

COMMERCIAL IN CONFIDENCE

Page: C-2 of 45
FUJ00001377

FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004

Appendix C System Management Version: 1.0

COMMERCIAL IN CONFIDENCE Date: 5/5/2000
0.3 Associated documents
[Reference Version Date Title ‘Source
ITD/ARC/029_ 0.4 42/11/99 [TMS Architecture Specification ICL Pathway
ITD/ARC/030 0.4 12/11/99 IOPS Architecture Specification ICL Pathway

0.4 Abbreviations and definitions

Abbreviation Definition

AlS Automated Interface Specification

BST British Summer Time

(CM (Configuration Management

cP Change Proposal, in ICL Pathway change management

process

CSV (Comma Separated Variable

DLL Dynamic Link Library

EPOSS Electronic Point of Sale System

OBC (Operational Business Change

OLA (Operating Level Agreement

OPS Office Platform Service

PinICL Record of a problem in ICL’s Problem tracking system

POCL Post Office Counters Limited

PVCS (Configuration management tool from Intersolve Inc.

RIPOSTE Retail Integrated Point Of Sale system in a Transaction
Environment: product from Escher that provides both the
infrastructure and the Desktop environment of the Horizon
system. The definitions in this manual refer to version 6 onwards

ISLA Service Level Agreement

[AN ide Area Network

0.5 Changes in this version

Version Changes

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-3 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
0.11 Review comments from version 0.10 have been incorporated.
0.12 Review comments from version 0.11 have been incorporated.
0.13 Review comments from version 0.12 have been incorporated.
0.14 Review comments from version 0.13 have been incorporated.
0.15 Review comments from version 0.14 have been incorporated.
0.16 Review comments from version 0.15 have been incorporated.
0.17 Review comments from version 0.16 have been incorporated.
0.18 [Review comments from version 0.17 have been incorporated.
1.0 Review comments from version 0.18 have been incorporated. Issue
for sign-off.
0.6 Changes expected
Changes
1. This document reflects the current implementation. The provided descriptions
and definitions may be subject to change control as determined by technical
and/or operational needs.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-4 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
0.7 Table of Contents
C.1 Introduction........ wT
C.2 Scope..... 1 8

C.2.1 System management environment...
C.2.2 Roles and responsibilities...

C.2.3 Life cycle responsibilities...

C.2.4 Reviews.
C.3 System Management and Package Distribution.
C.3.1 Overall software distribution processes.....

C.3.2 Software distribution and inventory managemen:
C.3.2.1 Sofware distribution monitoring.
C.3.2.2. Software inventory management.
C.3.2.3. Signing on distribution...

C.3.3 Reference Data distribution:
C.3.3.1 Reference Data interfaces.
C.3.3.2 Distribution of Reference Data...

C.3.3.3 POCL Reference Data change authorisation procedures..........
C.3.3.4 ICL Pathway Reference Data change authorisation procedure.
C.3.4 Software and Reference Data packaging rules....
C.3.4.1. Definition of software packaging rules.
C. 2 Rules for defining installation scripts.
C.3.4.3 Core and non-core software...
C.3.4.4 Software pilots. .
C.3.4.5 Sofiware and Reference Data interdependencies..

C.4 Application Configuration...

C.4.1 Hand-over Documentation.

C.4.2 Application Software Package Definitions...

C.4.3 Application Dependencies Definitions...
C.4.4 Application Security Model...
C.4.4.1 Access control at the outlet
C.4.4.2 Application access control.
C.4.5 Application Performance Model...
C.4.5.1 Performance characteristics of application.
C.4.5.2 Persistent object usage..
C.4.5.3. Message usage.
C.4.5.4 Use of indice:
C.4.5.5 Co-ordination across outlets.

C.4.6 Application System Description....

C.4.7 Application Support Documentation.. aeeeee
© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-5 of 45

ICL Pathway

FUJ00001377

FUJ00001377
Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C48

C.5 System Integration.

C.5.1
C.5.2
C.5.3
C.5.4

C.6 System Test...

C.6.1
C.6.2
C.6.3
C.6.4
C.6.5

C.7 Application Installed and Implemented...........

C71
C.7.2
C.7.3

C.8 System Supported......

C.8.1

C.8.2
C.8.3
C.8.4
C.8.5

C4.7.1. Event management...

C.4.7.2 Operational process control
C.4.7.3 Configuration management,

C.4.7.4 Error handling and diagnostic support
C4.7.5. Archiving and audit...

Application Training Documentation.

Build instructions.

Application software packages...........

Reference Data packages...

Fixes...

Test scripts.

User Operating Procedures.....
Performance monitoring......
Training course....

Fixes...

Counter Procedures..........
Training plan...

Fixes...

Application support.
C.8.1.1 Resource monitoring.
C.8.1.2 Version numbers...

Training.....

Performance monitoring......
Upgrades..........000.

Fixes...

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-6 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.1 Introduction

The Horizon system developed by ICL Pathway is a secure
environment incorporating a number of application services. It is
subject to Service Level Agreements (SLAs) governing the
performance and response of each of the services involved.

The development and implementation of a new application must not
cause any adverse effect on the existing service, or on the applications
within each service, thus compromising these Service Level
Agreements. Neither must it jeopardise the security of the Horizon
environment.

In addition, a number of Operating Level Agreements (OLAs) have
been established. These support SLAs by detailing operational
timetables and support arrangements.

Despite defining the APIs needed to develop new applications in this
environment (see main document), system management
considerations dictate that there may be an impact on the management
of the overall system. The system management characteristics of the
new application must therefore be established, as must its
performance characteristics.

The scope of the system management domain is described in section
C.2. The scope covers the components and documentation that must
be provided to support the system management life cycle. This life
cycle includes the integration, testing, release, and maintenance of
new applications.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-7 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
C.2 Scope

The existence of the Horizon systems management framework

imposes certain constraints on the behaviour of applications and the
way they are constructed. The phases system management life cycle
are shown in Figure C-1. For new applications the key elements are:

¢ Application Configuration where the application first enters the
system management domain. At this point all initial discussions and
reviews must have been completed and a definition of the
characteristics of the application established.

e System Integration where the application is integrated into the rest
of the Horizon system. At this point in the life cycle the key
constraints are the way the software is packaged, its dependencies
on other components, any security implications, its relationship to
existing Reference Data, and its need for new Reference Data.

e System Test where the application is tested in an end to end
system with the same configuration as in the live estate. Again,
software packaging is an important element as this is the first point
at which the live configuration is tested. Security is also important
as it may have operational implications. The way the system is
used in the outlet, as well as the support capability of the
application and its performance against predictions, are also tested
at this point.

e Application Installed and Implemented covers the installation of the
application throughout the live estate and the initiation of its use by
staff at each outlet. The co-ordination of the installation of the
software and Reference Data components and the user procedures
and the training schedule, are the key elements at this point
together with the ability to regress the implementation should errors
occur.

e System Supported where the support of the application, its
performance in the live estate, and the way any error fixes are
introduced into the estate become significant, as does the training
the users receive.

The way the various system management components fit into the
overall life cycle is described in section C.2.1. There are implications
for the roles and responsibilities of Third Parties, POCL, and ICL
Pathway. These are identified in section C.2.2 with the way they map
onto the life cycle covered in section C.2.3.

Irrespective of which organisation is responsible, a series of reviews
must be conducted to allow the resolution of any issues identified to be
embodied in the new applications, thus ensuring that the best solution

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-8 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

can be implemented to the benefit of the whole system. These reviews
are described in section C.2.4.

One of the main elements of system management is the distribution
process itself. Section C.3 covers the distribution of software and
Reference Data. The mechanisms used in each element of the life
cycle, and the documentation needed to support them, are covered in
the remaining sections. The implications of user support, training and
procedures are included where appropriate.

C.2.1 System management environment

The diagram below, Figure C-1, defines the components that must be
provided to support each phase of the System Management Life Cycle.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-9 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

Application Delivery

Hand-over Note
Application Software Package
an Definitions
Application Application Dependency Definitions
Build Instructions Configuration Reference Data Definitions
Application Software Packages Application Security Model
Reference Data Packages Application Performance Model
Fixes Application System Description
Application Support Documentation
SS ica Application Training Documentation
Test Scripts Se
Fixes Integration

User Operating Procedures
Performance Monitoring
Training Course

Fixes

Counter Procedures ivi
Training Plan activity
Fixes
-- Potential
- ) ICL Pathway
“~ Application activity

i Installed and
\\. Implemented

Application Support
Training
Performance Monitoring
Upgrades

Fixes

Figure C-1 System management environment

Figure C-1 identifies those activities within the system management life
cycle that ICL Pathway will undertake and those that ICL Pathway may
be subcontracted to undertake by a third party, see C.2.3.

If the third party supplier does not subcontract ICL Pathway to install
and implement its application, then the Risk Review, will be held to
determine the interactions with the ICL Pathway integrated systems.

C.2.2 Roles and responsibilities

The roles and responsibilities of those involved in the supply,
integration, testing, implementation and support of any new application

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-10 of 45

FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
must be agreed and the implications reviewed at the start of and
during the implementation of a new application.
The roles and responsibilities currently agreed with POCL are given in
section C.2.3 where the implications for system management are
identified. To support these roles and responsibilities, a number of
reviews are needed. The objectives for each review and the main
topics addressed are given in Table C-2, in section C.2.4.
ICL Pathway will assist POCL in identifying the roles and
responsibilities of all parties involved with third party software. The
areas where agreement is required include:
e Licensing arrangements
e Testing policy
e IPR
e Ongoing support and modification
e¢ Management of Change
e Liabilities, Cancellation and Termination
e Fault Management
¢ Service Introduction
C.2.3 Life cycle responsibilities
The table below sets out ICL Pathway’s view of its responsibilities
through the life cycle of a new third party application being introduced
on to OPS/TMS.
Activity ICL Pathway responsibility/role
Programme ICL Pathway will not be responsible. ICL Pathway will
Management provide suitable representation at Programme
meetings. This activity will contain the definition of
additional documentation or services to be provided by
ICL Pathway.
Business ICL Pathway will not be responsible, but wishes to be
Requirements involved from the earliest opportunity.
Systems Design ICL Pathway will be responsible for the overall design.

Application Design ICL Pathway will not be responsible.

Application ICL Pathway will not be responsible.
Development

Application Test and IICL Pathway will not be responsible. This is the unit
Integration and link testing of the application in isolation from other
applications.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-11 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

Systems Integration IICL Pathway will be responsible.

Systems Test ICL Pathway will be responsible. This is the end to end
testing of the system with the new application
integrated into the same configuration as the live
estate.

Type Approval ICL Pathway will be responsible for hardware and for
software type approval.

Business Acceptance IICL Pathway will not be responsible.

Manufacture For hardware ICL Pathway may or may not be
subcontracted by the supplier.

Distribution For hardware ICL Pathway may or may not be
subcontracted by the supplier.

Installation and For hardware ICL Pathway may or may not be

Implementation subcontracted by the supplier.

For software ICL Pathway may or may not be
subcontracted by the Supplier.

Support ICL Pathway may or may not be subcontracted by the

Supplier. All software changes must be integrated and
fully validated by ICL Pathway.

(Support means software fault detection and
rectification.)

Maintenance For hardware ICL Pathway may or may not be
subcontracted by the third party supplier.

(Maintenance means hardware maintenance.)

Training ICL Pathway may or may not be subcontracted by the
Supplier.

Help Desk support —_IICL Pathway may or may not be subcontracted by the
Supplier.

Service Reporting ICL Pathway may be responsible for reporting but will
not take Service Level Agreement responsibility for the
application.

This means ICL Pathway may be responsible for
producing reports of performance against SLAs.

Invoicing ICL Pathway may provide information on which an
invoice can be based, but will not be responsible for
invoicing or reconciliation.

This means ICL Pathway may provide information in
the form of reports from which invoices may be
prepared.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-12 of 45

FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

Table C-1 Activities and responsibilities
C.2.4 Reviews

System Management starts after the software and data that comprise
the application are delivered as packaged components. Before any
application can be introduced into the system management
environment, a number of reviews must have been conducted:

e Dependencies Review.
« Performance Review.
e Process Review

« Risk Review.

e Security Review (see also Appendix B, Cryptography and Key
Management).

e SmartMan Review (for smart card applications) (see also Appendix
A, SmartMan Interfaces).

e System Management Review.

Note that these reviews may need to be repeated after the application
has been delivered.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-13 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
(Review Objective and Main Topics

Dependency Establish the dependencies of the new application on the other partsI
of the Horizon system. The main topics are:

e Synchronisations.
¢ Component lists.

e Dependencies.

Performance IDefine the expected performance of the system so that the
Performance Model can be established and to support its validation
during System Test. The main topics are:

e Rate of change of software.

e Rate of change of Reference Data.

e Use of the General Acknowledgement Agent.
e Expiry periods.

e Volumes of transactions, Reference Data changes and resource
usage.

e Frequency of events.
e Use of persistent objects.

e Use of indices.

Process Establish and document the processes carried out by the clerk at the
counter. The main topics are:

« Counter Procedures.
e Help provided to the counter clerk.

Risk Establish and quantify the risks where ICL Pathway is not being
contracted to install and implement a new application. The main
topics are:

«Interfaces.
e Interactions.

e Impacts.

Security Establish and quantify the impact of existing security facilities, and
the scope of any changes required. The main topics are:

« Access control.
« Domains.
e Security log.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-14 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
SmartMan Identify and quantify the impact of introducing a new smart card

application, including:

e Smart card technology to be used, whether it can be supported.
e Firmware enhancement (of the card reader/encoder).

e Impact on existing SmartMan implementation.

e Recovery conditions.

« Use of menus.

System Identify all components and the way they will be integrated, tested
Management and distributed, together with the support needed during System
Integration, System Testing and live operation. The main topics are:

e Installation, implementation and support.

e Restarts, and manual intervention.

e Software distribution.

« Audit Service.

e Additional support from the Horizon Help Desk.

e Separate Help Desk for the new application.

Table C-2 Review objectives

The result of these reviews are agreed deliverables, interfaces, and
documentation that provides the level of understanding that allows the
application to be configured and delivered into the System Integration
testing. After the integrated system has been System Tested, the
application can be released into the live environment and become part
of the supported system.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-15 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.3 System Management and Package Distribution

Software has to be packaged so that it can be distributed effectively in
the distributed environment of 20,000 Post Office outlets.

The primary software distribution tool used in the Horizon solution is
Tivoli. Application Products and Items and their design documentation
must be registered in PVCS. The registered application Product and
Items can then be converted into Tivoli file packages and installed
(and if necessary un-installed) by Tivoli in unattended mode. The
overall software distribution process is described in section C.3.1, with
the way each package is installed covered in section C.3.2. The way
the installation is monitoring is also described in section C.3.2, as is
the way the inventory is maintained.

Some Reference Data is distributed through the Reference Data
Management Centre and the rest as work packages. These
distinctions are described in section C.3.3.

There are also rules that have to be followed when creating packages
for distribution. These are covered in section C.3.4.

C.3.1 Overall software distribution processes

The same packaging is used during integration, testing, and the
maintenance of the live estate. The scope of the Software Distribution
process is as follows.

ICL Pathway Fett
ou

oe a

Repository

ICL Pathway

Campuses and I
Outlets
1G. Patiway CS ee :
Software Depot ii serang si)
Test Rigs : Wodows NT Platforms Sequent Prtforms

Figure C-2 - Software Distribution

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-16 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

Software is lodged in the ICL Pathway Configuration Management
(CM) Software Repository. It must have an identifiable Name and
Version Number, see section C.4.1 and section 8 in the main
document. These are carried with it throughout the distribution
process.

When it is ready for release, it is transmitted via the CM Signing
Server to either the ICL Pathway Software Depot for testing at
Bracknell, or the Outsourcing Software Depot for distribution to the live
estate. In both cases, it is first verified as being ready for release. The
purpose of the distribution to Bracknell is to verify that the Tivoli
distribution packages can be applied successfully, and that the
package can be regressed if necessary.

The CM Signing Server ensures that all packages are signed so that
any corruption can be detected on arrival at the outlet. The base
Riposte software is separately signed to ensure that only the correct
OPS environment is loaded at each counter position, see section
C.3.2.3.

C.3.2 Software distribution and inventory management

C.3.2.1 Software distribution monitoring

Tivoli Distribution Monitoring runs on each client platform and monitors
the operation of the installation process. It returns status information to
the inventory, which records the system status and the presence of the
installed files.

C.3.2.2 Software inventory management

The inventory uses a central Oracle database within the systems
management function to maintain details of the platforms, software and
applications asset base in the installed OPS. It includes a Web
browsing capability that enables support staff in a wide variety of
locations to examine the status of distributions, the software state of
any managed node, and any problems that have occurred with a
distribution. The database is also used by other system components,
such as the software distribution service.

The database is hierarchically structured and is extensible to cope with
any new hardware or software objects or attributes deployed in
response to future services or applications needs. It is updated
automatically when any new equipment is installed in an outlet, or
when new or updated software is distributed and installed.

Any changes needed to support the introduction of a new application
must be raised at the System Management Review.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-17 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
C.3.2.2.1 Inventory signatures

For security and integrity purposes, signatures are applied to platforms
and products:

e Platforms Each counter has a signature identifying the base build
software. This signature is also held in the inventory for that
counter on the central management system.

e Products Each change to a product on a particular counter has an
unique signature held on the counter itself and also in the inventory
for that counter on the central management system.

These signatures are maintained by the ICL Pathway distribution
mechanism. The signatures name one or more constituent files. All of
these signatures on a counter can be remotely inspected from the ICL
Pathway central management system.

C.3.2.3 Signing on distribution

All application software packages are signed for distribution to outlets.
Software that is distributed is signed by the CM Signing Server (see
C.3.1). Additionally, those EXEs, DLLs, and files associated with the
Riposte Desktop and message server and any security related
software are individually signed and the signatures are checked each
time the code or file is loaded. This ensures that the correct
environment for applications is established each time that the Desktop
is loaded. This signing is in addition to the signing of work packages
for Tivoli distribution purposes.

C.3.3 Reference Data distribution

Reference Data is used throughout the Horizon system for a variety of
purposes. It is used to support the business processes in the outlets in
the live estate, client operations as well as determining the products
and pricing that support the generation of Cash Account reports at
each outlet. It also defines the interpretation of the use of counter
peripheral, as well as defining the dialogue that controls the actions
carried out by clerks for individual products and services.

Reference Data is fundamental to the accounting integrity of the
system. The make up of the cash account is defined through
Reference Data, for instance. Reference Data also plays a
fundamental role in roll out, for example during migration.

C.3.3.1 Reference Data interfaces

For each of the five types of Reference Data used by an application a
separate Reference Data interface is used, see section 3.1.2, Data
interface types, in the main document.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-18 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.3.3.1.1 Business Type A

Type A is POCL Reference Data that conforms to the POCL
Automated Interface Specification (AIS).

C.3.3.1.2 Business Type B

Type B data originates from POCL but is not included in the automatic
AIS interface. This data is transferred in different formats (e.g.
spreadsheets, CSV files, text documents as part of the Operational
Business Change process).

C.3.3.1.3 Implementation Type C

This is implementation data generated entirely by ICL Pathway, often
at the behest of POCL.

C.3.3.1.4 Implementation Type D

This data is also known as ‘Build Objects’ in that it is Reference Data
generated by ICL Pathway concerning the underlying Horizon build. It
contains Riposte Objects and Training data (to enable Training Mode
to be entered from the Live Desktop).

C.3.3.1.5 Client Reference Data

This interface is implemented by the application concerned. An
example of this kind of data is tariff data.

C.3.3.2 Distribution of Reference Data

C.3.3.2.1 Delivery of Business Reference Data

The ICL Pathway Reference Data Management Centre (RDMC) holds
POCL supplied Business, and some Implementation, Reference Data
in an Oracle database. This database acts as the source for the
distribution process. The Reference Data Distribution Server (RDDS)
creates agent views of the Reference Data by pre-processing the
Reference Data into attribute grammar form. Such Reference Data is
held in (attribute grammar) ‘Collections’, each Collection containing a
set of related Objects (themselves decomposing into Attributes).
Figure C-3 illustrates the distribution mechanism.

Business Reference Data is held as temporal persistent objects.
Temporal objects have a start date and (optional) end date. Changes
to temporal objects are recorded by end-dating one instance of that
object, and adding a new occurrence with new start and end dates.

Where any item within an object has changed, the pre-processing
stage creates a new version containing all the items within that object
that apply from the date that the change is valid (allocating this a new
version number).

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-19 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
POCL Ref. Business Type B
Data v
Business Type A Business Type B ICL Pathway Reference
Data Management
I¢—______I
RDMC
Implementation
v Type C

RDDS

Agent -

Implementation
Type D
Correspondence

Server

Counters

Figure C-3 Reference Data distribution
C.3.3.2.2 Delivery of Implementation Reference Data

Implementation Reference Data, that is Type C and Type D data, is
held as persistent objects in collections, but the data is not temporal in
that it remains extant until either replaced or deleted.

Reference Data used by new applications will be provided as a set of
collections using the naming conventions and collection names
identified in sections 8 and 9 of the main document. The only
exception to this rule is the Reference Data associated with Access
Control, (see section C.5).

The arrival of specific Reference Data at an outlet can, if necessary,
be monitored by use of the General Acknowledgement Agent, see
section 7.2.1, General Acknowledgement Agent, of the main document
for details.

Because of the performance implications of such an approach, any
usage would need to be quantified during the Performance Review.

C.3.3.3 POCL Reference Data change authorisation procedures

The complete list of all Reference Data changes allowed for Types A,
B, and C is set out in the Reference Data Change Catalogue
(CS/IFS/001). This categorises changes into three types: basic
(planned), advanced (planned) and unplanned.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-20 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.3.3.3.1 Basic Changes

Basic Changes are those that ICL Pathway can support without prior
notice and only require basic control and release activity by ICL
Pathway. The data contained in these is referred to as ‘Class1’ data,
for example a change to an outlet’s telephone number or change in
product price

All Basic Changes are supplied by POCL as Type A data and are
released without reference to the Operational Business Change (OBC)
process that controls releases of changes into the live estate.

C.3.3.3.2 Advanced Changes

Advanced Changes are those that need additional activity by ICL
Pathway before the change can be implemented.

Advanced Changes are always authorised via the OBC process, and
are supplied via Type A, B, or C Reference Data, or could be
implemented via Type D Reference Data.

C.3.3.3.3 Product Changes

A Product Change is the change needed to the definition of a product.
It is always supported through Reference Data and may be either a
Basic or an Advanced Change. Table C-3 identifies the relationships
between categories of change and types of Reference Data involved.
Scales Reference Data is supplied via Type B Reference Data, and is
considered as product data.

\Categories Type Examples

Basic Change - Pure Type A Change clerk instructions (and possibly
changing non-core product availability).

Basic Change -—High ITypeA Product price, product names,

Risk revaluation, method of payment, change
whether voidable/reversible, change
min/max quantities/values.

Advanced Simple Type A plus after ICease non-value product, Non core
Change the event product becomes core, Change to use of
changes additional fields.

Advanced Standard Type A and Type IChange to ability to print receipt, change
Change B (also MIS value to non-value stock, picklist, remove
change) AP Client, calendar.
Advanced Complex Type C New product, change scales tariff, screen
Change layout, change accounting node, change
product core to non-core.
AP Changes AP Client Take INew AP Client, client name change,

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-21 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
On cease AP Client, product or token.
Fastrack — Basic Type A Price of non-value or re-valuable value
Express stock, change product name.
Fastrack — Migration Type B Correcting Reference Data to show non-
Special core products at outlets.
Fastrack — Tight Any Price changes.
Timescales
Fastrack — Error Any Any.
Correction
Table C-3 Product changes
C.3.3.4 ICL Pathway Reference Data change authorisation
procedure
For Types C and D Reference Data, changed or new data is released
via work packages and follow the normal OBC procedure for
authorisation of the work package into the live estate.
C.3.4 Software and Reference Data packaging rules

Since software is distributed in units called file packages, these
packages must be analysed, scripted, and tested before they are
handed over. They must include the development of a regression
capability. A package in this context can represent a single file, a set of
related files of an entire application. Whether or not the package
requires a system reboot must be defined. Each file is ‘committed’ in its
own right, and so must be self-consistent.

Software Distribution operates in a hierarchical manner. A package
can be installed on any Windows NT server, and then distributed from
that server to the network resources it serves. Packages are passed
first to a local Tivoli server. This distributes it to the target environment
(for example outlets), using the distribution-related hardware and
network inventory to determine the distribution targets and network
routes. The distribution process takes account of the network
bandwidth available.

Within an outlet, the Gateway PC forwards the package to the other
counter PCs. This considerably reduces the potential load on the
network. The distribution can be fine-tuned to exploit a designated
amount of bandwidth at each level of the hierarchy.

The Tivoli Commit operation is used on-line, to run the installation
routine on the outlet counter PC in unattended mode. The package is
re-verified before installation takes place. This routine creates a
regression capability, for example taking before-looks of the NT
registry, copies of replaced DLLs and so on. Installation is unattended,

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-22 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

to avoid relying on (non-existent) technical support at the target site. In
order to support this process the rules defined in sections C.3.4.1 and
C.3.4.2 must be followed.

Software changes are introduced during the early morning, when the
Riposte Desktop and services are shutdown. The shutdown and restart
are controlled by Tivoli. Because re-booting counter PCs requires
intervention from the Post Master in terms of the Post Master Logon
Process (POLO) (described in section 6.3.3, Reboot, of the main
document) such re-boots are kept to a minimum.

The rate of change of software can have performance implications,
and so the rate at which it is predicted to change must be established
during the Performance Review so that it can be modelled and
monitored.

C.3.4.1 Definition of software packaging rules

The definition of each software product or product upgrade must be
defined, in terms of:

¢ Its constituent files.

e Scripts to perform the installation (and removal) of the product, see
section C.3.4.2 for the rules governing scripts.

e Criteria by which it can be asserted that a software product is
installed (sometimes called a software signature).

e Anaming scheme for identifying the product.

The Hand-over documentation required is described in section C.4.1
and the Application Software Package Definition in section C.4.2.

Packages must be design to ensure that they can be installed
remotely, without any impact on other packages. Large package cause
distribution issues. Accordingly, the target maximum size is 0.5Mb.

All changes to applications must be supplied as deltas and must
provide some persistent signature as to their existence and co-exist
with the target platform environment.

If there are any issues with any packaging rules, they must be raised
during the System Management Review.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-23 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
C.3.4.2 Rules for defining installation scripts

Each script provided must:

e Allow the software product or product upgrade to be distributed to
one or more end systems.

e Becapable of activation on the destination system(s), including
supporting the execution of any necessary data translation
functions within the scripts. All existing data structures must
continue to be supported during the migration period, when more
than one version of a package can exist within the estate at the
same time.

e Be capable of supporting reversion to a previous version of a
software product.

e Support the deletion of a software product on one or more end
systems.

C.3.4.3 Core and non-core software

The strategy for distribution of software is to distribute all software to
the outlets, and to use Implementation Reference Data to control its
use.

C.3.4.4 Software pilots

It is possible for pilot purposes to distribute software to a selected
number of outlets.

Should a software pilot be a requirement then the issue must be raised
during the System Management Review.

C.3.4.5 Software and Reference Data interdependencies

Business Reference Data is temporal in nature and so can be
distributed in advance of the new version of software that uses it.
Implementation Reference Data is application specific and so can be
distributed in advance of the new application that uses it.

Not all counters will receive the new version of application software at
the same time, so any dependencies on the order of distribution of
counter and agent packages must be included in the dependencies
definitions, see section C.4.3.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-24 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.4 Application Configuration

The Application Configuration is the first phase at which each
application enters the system management domain. At this point, all
initial discussions and reviews need to have been completed and a
definition of the characteristics of the application established. The
ease with which an application can be implemented, distributed, and
supported depends on the quality of the documentation that describes
its behaviour.

The documentation that needs to be provided includes:

e Hand-over Note that defines the packages being handed over, see
section C.4.1.

e Application Software Package Definitions that defines the
components being handed over in each package, see section
C.4.2.

¢ Definition of all application dependencies, see section C.4.3.

e Application Security Model, see section C.4.4.

e Application Performance Model, see section C.4.5.

e Description of the application being delivered, see section C.4.6.

e Documentation needed to support the application, see section
C.4.7.

e Training documentation, see section 0.

C.4.1 Hand-over Documentation

The supplier of a new application must provide a Hand-over Note that
allows the hand-over to be recorded and defines all the Products and
Items lodged in Configuration Management that relate to the
application being handed over.

The table below summarises the items to be included:
Hand-over Note

tem (Description

‘Supplier [The name of the supplier making the hand-over.

Hand-over Identifier Uniquely identifies the particular hand-over.

Hand-over Description of the contents of the deliverable.

Description

Owner Name of the person responsible for the hand-
over.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-25 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
Physical Media to be labelled with Supplier ID, Hand-over
imedia/Delivery Description, sequence number of the media. The
location type of media that is acceptable is diskette or CD.
ROM.
(Components Details of all components in the deliverable.
Table C-4 Hand-over Note contents
These definitions, for which there is a Package Definition, see section
C.4.2, must include all software and Reference Data Work Packages,
and Build and Installation Scripts. The scope and content of each
package must be agreed at the System Management Review.
C.4.2 Application Software Package Definitions
Each software package must have a unique identity and name each
component within a package. The documentation used to define
building the application is summarised in Table C-5, below. It includes
the scripts that support the build, which must be supplied as separate
work package definitions.
[tem \Description/sub-item
Name Name by which the item will be known.
\Type Optional, but if supplied will form part of the
physical name.
ersion ersion by which the item will be known.
Components All the components that are used to build the
deliverable.
‘Component Name \As specified to CM, but
ithout a suffix.
(Component Version _IThe version to be used in
ithe build.
‘Component Type (Any file suffix.)
Related Change List of all Change Proposals (CPs) incorporated.
Proposals The definition of the content of numbered
changes needs to be maintained for integration,
system test and support purposes.
Related test List of all faults (PinICLs) for which a solution hasI
jobservations been provided. The definition of all faults
resolved needs to be maintained for integration,
system test and support purposes.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-26 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

Products serviced List of products that are to be serviced by the
(deliverable. Installation details are required for
leach entry.

Table C-5 Contents of application software package build instructions
C.4.3 Application Dependencies Definitions

Software developed to operate in the Horizon environment has impact
across horizontal platforms and vertical ‘slices’ as described in the
OPS and TMS Architecture documents. To be able to achieve such
system integration all dependencies both horizontal and vertical must
be identified.

The dependencies may be on other OPS or TMS components and
between application components themselves. The main dependencies
that must be identified are:

e OPS hardware including configurations, connectivity and usage is
described in section 8, Physical Architecture, in the TMS
Architecture document.

¢ OPS software including Windows NT, Riposte Desktop and
software libraries and registry settings.

e TMS hardware including configurations and usage at both counters
and correspondence server levels as described in section 8,
Physical Architecture in the TMS Architecture document.

e TMS software including Riposte Message Server and software
libraries version and registry settings.

e Interfaces to and from the TMS layer in terms of physical
characteristics and recovery mechanisms.

¢ Interfaces to the security software used in the OPS and TMS
domains.

There are also potentially dependencies on existing services,
including:

« Audit service.

« Archive service.

e Help Desk.

e Reference Data Management services.

All these dependencies must be identified, and the impact analysed
and quantified during the Dependency Review process.

The definitions that result from the Dependency Review must identify
all document, software and hardware dependencies, as illustrated in
Table C-6.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-27 of 45

FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
tem Description
Input documents ‘A complete list of the documents that were used
for the production of this product.
Software Definition of other software needed to support
dependencies the integration of the deliverable.
Hardware Definition of any hardware required to support
dependencies ithe integration of the deliverable.

‘Service dependenciesIDefinition any processes or NT services that
may impact successful start-up or shut-down.

Data dependencies Definition of all Reference Data objects used.

Table C-6 Dependencies definitions

Should a new application require additional support from the Horizon
System Help Desk, then the impact of this support must be discussed
with ICL Pathway during the System Management Review.

If a separate Help Desk is established, then any feed of information
from the ICL Pathway system must be agreed in advance during the
System Management Review.

Should a new application need to support HTML Help screens as well
as the standard bubble Help, then the impact on the Performance
Model must be discussed during the Performance Review and the
support implication during the System Management Review.

C.4.4 Application Security Model

To ensure that access is controlled, controls are applied to the
application functionality. The Security Model that applies to a specific
application must be covered during the Security Review where:

e Any cryptographic and key management requirements must be
identified, see Appendix B, Cryptography and Key Management.

e Any access control rules at the counter must be established, see
section * .

e Any access control rules that the application needs to police must
be established, see section C.4.4.2.

C.4.4.1 Access control at the outlet

Access Control is applied at a resource level. It means verifying that a

particular user has the right to access a given resource in a particular

way. Access control is defined in terms of Roles, each of which defines
a number of functions that a user can use. Effective access control

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-28 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

depends on having a clear definition of the roles and responsibilities of
everyone who needs some form of access to the system. A user may
be allowed to use several different roles, which may or may not
provide overlapping sets of functions. Authorisation is the process of
determining whether a user has permission to use a particular role,
whether this is formally defined or not.

C.4.4.1.1 Start-up process

Control of access to the facilities within the OPS on each counter PC
within an outlet is by means of the Post Master Logon (POLO)
process. A smart card is used to authenticate that the Post Master is
allowed to access to the counter PC.

C.4.4.1.2 Log on process

After the POLO process has been completed, users logon to the
Desktop in one of the roles defined for an outlet. The access control
mechanism for applications within the OPS is provided by facilities in
the Riposte Desktop software.

C.4.4.1.3 User roles

The roles that have been agreed with POCL to provide access control
to the Desktop at an outlet are:

e Post Office Roles, including Post Office Manager (Manager and
Supervisor), Counter Clerks, (Clerk) and Auditor, (Auditor).

e Operational Roles, which provide the means to control the Horizon
systems during normal running (Emergency Manager).

¢ Support Roles, such as engineers (Engineer) and support
(Support, Migrate, Setup).

The access rights of each type of use are defined in persistent objects.
Users with Post Office roles have passwords that have to be changed
regularly. All other users have to use a ‘one-time’ password that
requires a call to the Horizon Help Desk.

Any implications of these rules on a new application must be raised
during the Security Review.

C.4.4.2 Application access control

The Desktop itself is a hierarchic menu driven system, and this feature
is exploited to allow only certain roles access to certain parts of the
menu structure. This is achieved by establishing access rights to
buttons within the hierarchy.

Applications are accessed from a particular level in the hierarchy, and
are responsible for checking that they are being accessed only by
those roles authorised so to do.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-29 of 45

FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

For each application, the roles by which it can be used must be
defined, and in turn, the application must check that it has been called
by an allowable role.

The definition of these details must be provided as part of the Hand-
over Note and discussed during the Security Review process so that it
can be incorporated in the Menu Hierarchy document.

C.4.5 Application Performance Model

A new application will cause some effect on the volume of transactions
and hence the performance of the Horizon system.

The Performance Model for an application must be established during
the Performance Review. The topics that must be addressed and
documented in the model are:

e Application performance characteristics, from the point of view of
the volumetric requirements as identified in section C.4.5.1.

e Usage of persistent objects, bearing in mind the issues identified in
section C.4.5.2.

e Usage of messages, bearing in mind the issues identified in section
C.4.5.3.

e Usage of indices, bearing in mind the issues identified in section
C.4.5.4.

e Impact of the use of any online transactions, with reference to the
potential performance impact identified in section C.4.5.5.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-30 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
C.4.5.1 Performance characteristics of application

For a new application, the volumetric characteristics must be
established during the Performance review so that they can be
included in the Performance Model prior to the implementation of the
application. The model can then be monitored when the application is
system tested and when it is released in the live estate, see section
C.6.3.

The capacity of each component must be modelled by platform as well
as the predicted volume of data passing across the interfaces between
components such as agents, WAN lines and the application. During
System Tests and live running, the capacities are tested against the
declared volumes for these components.

The following are required for each Business Level Transaction in the
new application:

1. The number of transactions for each hour of the week (i.e. the
values for 24 hours a day, 7 days a week) in the system in a typical
week.

2. The number of transactions for each hour of the week in the system
in a seasonal peak.

3. The number of transactions for each hour of the week in the pre-
Christmas peak.

4. The distribution of transactions versus the size of outlets.
5. When seasonal peaks, other than Christmas, will occur.

6. The details of the transaction including Riposte messages
produced, any interactions with the data centre, response time
requirements etc.

7. Resource usage (processor, disc capacity, disc i/o, memory, LAN
usage etc) for every platform that the application uses (e.g.
counter, agent, host).

8. The volume and frequency of Reference Data changes.

If any part of the application runs in the data centre (e.g. an agent) the
following performance statistics need to be established:

e The volume of transactions that the application will support.

e The number and type of RPC calls made against other components
of the system (e.g. RPC calls by an agent to the correspondence
server).

e The elapsed time for batch processes.

These volumes must be formally agreed both in the Performance
Model and as a contract change during the Performance Review

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-31 of 45

FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

because they will be the basis of any cost that ICL Pathway incurs
should a new application require it to increase the capacity of the
system.

C.4.5.2 Persistent object usage

By their very nature, persistent objects should not change very often.
The performance implications at the Correspondence Server level are
severe if they are created or changed frequently.

Any usage at all of persistent objects, except for that defined in section
5 of the main document, must be agreed with ICL Pathway during the
Performance Review and documented as part of the Performance
Model. Wherever possible an alternative approach using transient
messages will be proposed.

C.4.5.3 Message usage

Because of the volume of messages currently processed by the
Horizon system, care must be taken to create the minimum number of
additional messages consistent with the requirement of the application.
Message expiry periods should be kept as short as possible consistent
with the business requirements and potential for communications
outages.

There is a default expiry period that will be used if the application does
not set an explicit value. ICL Pathway reserves the right for operational
reasons to change the default expiry period without notice for
performance purposes, but will not reduce it such that the requirement
to hold the data for two complete previous POCL Outlet Accounting
Periods (normally seven days each) is compromised. So, it is important
that Business specific values are used wherever possible.

The volume and size of all messages must be discussed during the
Performance Review so that they can be incorporated into the
Performance Model. The content and structure of all messages that
flow across interfaces must be discussed during the System
Management Review and documented in the Application System
Description.

The expiry period for each message created by an application must be
agreed with ICL Pathway during the Performance Review so that its
anticipated effect can be established using the Performance Model,
and its actual effect measured on the live estate.

C.4.5.4 Use of indices

Indices are only to be used for truly random access in performance
critical operations. The way indices are created is discussed in section
4 of the TMS Architecture Specification.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-32 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.4.5.4.1 Correspondence servers indices

Because of the impact on performance caused by the use of indices at
this layer, the only indices that may be used at the Correspondence
Server layer are those created by ICL Pathway.

This results in the following rules being applied to indices created and
used on the correspondence servers:

e Applications must not create indexed messages or persistent
objects.

e Indices must not be used to find all objects of a particular type.
(The Riposte Scan function is very efficient for this purpose.)

e Applications must not destroy any indices.
C.4.5.4.2 Counter indices

The following rules apply to indices created and used on counters:

e Applications must only create indices if absolutely needed for
random access.

e Indices must not be used to find all objects of a particular type.
(The Riposte Scan function is very efficient for this purpose.)

e Applications must not destroy any indices.

The indices created and used by the application must be established
during the Performance Review so that they can be included in the
Performance Model. This is of particular importance if new indices are
required by counter applications, as a check must be made to ensure
that they have no effect on the performance of other applications.

C.4.5.5 Co-ordination across outlets

Due to the nature of the distributed system, it is possible for a new
application to introduce additional implicit synchronisations across
multiple outlets. These synchronisations may be acceptable if they do
not impact the data centre.

An example of a scenario that would have a significant performance
impact would be if a new application did additional processing at the
end of the day at a different time from the existing End of Day
processing and required an on-line transaction to be enacted at this
time.

Any elements of the design of a new application that have implicit
synchronisation implications must be raised during the Performance
Review so that the impact can be evaluated.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-33 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.4.6 Application System Description

The Application System Description defines the overall system design
and is a key document in terms of its impact on the successful
management of the system when implemented. This document must
ensure that:

e Allinterfaces and the data that flows across them are identified.
¢ All interactions with the clerk are identified.

e All operational aspects are identified, including use of a Help Desk
to resolve clerk concerns.

¢ Business Process flows are identified to enable Counter
Procedures to be created.

The contents of this document must be discussed during the System
Management Review.

Should the application use a smart card, then any issue in this area
must be raised and resolved during the SmartMan Review, see
Appendix A, SmartMan Interface.

C.4.7 Application Support Documentation

The Application System Description defines the overall design of each
new application. The Application Support Document defines the
operational aspects of the new system.

There are a number of aspects that must be addressed in this support
documentation . These include:

e Any event recorded by the application must be agreed and
documented, see section C.4.7.1.

e Operational aspects such as how to start and stop the application
must be agreed and documented, see section C.4.7.2.

e Configuration management aspects of the operational system must
be agreed and documented, see section C.4.7.3.

e Error handling and diagnostic support must be agreed and
documented, see section C.4.7.4.

e Archiving and audit aspects, see section C.4.7.5.

These operational aspects must be raised during the System
Management Review.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-34 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
C.4.7.1 Event management

Event management is used to record information about the behaviour
of the applications present in each outlet. It is the responsibility of each
application to record in the NT Application Event Log file, those events
that can have a major impact on the use of the application.

For each event the severity of the event must be defined, as must the
action to be taken once the event has occurred. Because of the
distributed nature of the solution, ICL Pathway has a filter mechanism
that allows specific events to be forwarded to the centre for action.
Since all support is remote, the application must be self-managing
wherever possible, and where this is not possible there must be an
agreed support procedure.

There must be agreement with ICL Pathway during the System
Management Review on the impact and resolution of all events and
each event must be documented in the Application Support document.

There are rules that each application must follow with regard to the
use of events. These are:

e Use only the NT Application event log.
e Identify all source names used.

e Document all events, their severity, and suggested recovery
actions. These include system errors, application errors, time-outs,
and thresholds exceeded.

e Ensure that the application correctly uses NT event types (i.e.
errors, warnings, and information). Events must include a unique
indication of the application's name and version.

e Document security related events or security enforcing events, i.e.
that they affect or are reporting on the security integrity of the
application and platform.

e Indication of the likely event volumes for normal running.

e Avoid floods of events through duplicates. The applications should
treat these as inherent application state changes, and fall back to
periodic error reporting on the state.

Any potential use of the NT Security Event Log must be agreed at the
Security Review.

C.4.7.2 Operational process control

If the application is a standalone process outside the Desktop, the
Application Support document must:

e Document if it may be safely started and stopped. This may be by
an API or may involve the use of the specific executables. If an API

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-35 of 45

FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

is provided then it must be synchronous and atomic, i.e. only return
when all, or no, processes are started or stopped. The definition of
each must be included in the components identified in the
application configuration and dependencies documentation. Any
peripheral interactions must also be agreed and documented.

e Document any dependencies to other processes or NT services
that may affect their successful start up or shut down. These must
also be included in the dependencies documentation.

e Document whether they can be (safely) restarted automatically on
failure.

If the application is implemented as an NT service, the Application
Support Document must:

e¢ Document all service names used.

e Document any dependencies to other processes or NT services
that may affect their successful start up, or shut down.

e Document whether they can be (safely) restarted automatically on
failure.

If a service cannot be safely restarted, then the implication of this and
any manual intervention required must be raised at the System
Management Review.

C.4.7.3 Configuration management

For any configuration data that is not held as persistent objects in the
message store, the Applications Support document must:

e Document any generic configurations, the APIs for setting the
parameters and the underlying footprint on the NT registry, if used
as the repository for these parameters.

e Document when changes to these parameters take effect. This
must either be when the parameters are updated or by invoking
another API. The necessity to restart the application is strongly
deprecated.

e Document any dependencies on the configuration of other
applications/NT subsystems and indicate the synchronisation of
these dependencies.

The same documentation must be provided for all configuration items
within TMS and OPS.

The use of all parameter settings must be discussed as part of the
System Management Review.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-36 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
C.4.7.4 Error handling and diagnostic support

All counters are remotely managed and have no onsite IT experienced
support yet must be continuously available. The application design
needs to be aware of this and obey engineering principles that support
this environment. They include:

e The application must never silently fail, that is wherever possible it
must record the reason for failure, even though the failure may not
be made visible to the clerk, for fraud prevention reasons.

e The application must provide diagnostic facilities that can be
remotely managed. The preferred model is:

= An API is provided to switch on/off specific diagnostics. Such
changes shall be immediately actioned by the application, or a
separate API is provided to make the application recognise the
new values. Applications must not need to be restarted.

= Such diagnostics shall go to one or more diagnostic files. All
files must be cyclic, whether disc or memory resident.

= The diagnostic files shall be in English text format, or if encoded
in other ways (e.g. binary, local language) then there must be
facilities to transform the data into English text format at source.

e Applications must never rely on an operating system reboot or
application restart to free dynamic resources such as memory,
semaphores, temporary disc storage.

e Applications must successfully manage transitions across time
changes, such as in BST. At a minimum their behaviour shall be
documented in conditions where the time changes (e.g. if the
application relies on any time based recovery) and they must never
fail.

e Applications must provide recovery from all conditions. Ideally, this
is always automated, but if not, documentation must be provided to
describe the recovery action. Recovery conditions must cover an
uncontrolled system shut down (e.g. power loss) and therefore an
application must design out (or minimise) exposure to critical
sections in the code.

The principle being use to deal with error recovery and diagnostic
support must be discussed and agreed during the System
Management Review, and the outcome documented in the Application
Support Document.

C.4.7.5 Archiving and audit

Every message within the message store has an expiry period as one
of its attributes. This is the number of days that the message will stay

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-37 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

on-line in the message store. It may be set explicitly by the application
that creates the message, or it may default to the system default that is
set as a Riposte configuration parameter.

An expired message cannot be retrieved from Riposte even if it has not
yet been archived. (See the main body of the document on the
temporal nature of messages.)

Archived messages are deleted at the outlet and archived by the
Archive Service at the Correspondence Server level to other media
and then deleted.

Should any data created by an application be required for audit
purposes then this must be discussed with ICL Pathway during the
System Management Review to establish the effect on the existing
Audit Service.

The use made by the application of archive and audit facilities must be
included in the Application System Description and any operational
implications in the Application Support Document.

C.4.8 Application Training Documentation

For each application a training course may need to be developed and
training material must be created for the Training Mode capability
provided to all users in an outlet.

These two requirements must be discussed during the System
Management Review and agreement reached on the documentation
required.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-38 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.5 System Integration

At the point in the life cycle where a new application is integrated into
the rest of the Horizon system, there are a number of key constraints
that must be addressed. They are:

e The way the software is packaged in terms of its components,
dependencies and any security implications.

e Its relationship to existing Reference Data.
e Its need for new Reference Data.

These will have all been documented at the Application Configuration
stage. This is the first point in the System Management life cycle when
errors can be raised on the delivered software and Reference Data.
These errors will be registered and progressed through the ICL fault
management system PinICL.

The deliverables at this stage are therefore:

e Build instructions that define the parameters needed to support the
ICL Pathway platform build process.

e Application software itself.
e Any Reference Data specific to the application.
e For any fix for an error identified, specific details of the change.

C.5.1 Build instructions

ICL Pathway has an automated method for building platforms for test
purposes. The definitions provided for a new application must be
sufficient to enable this process to be extended to include the new
application.

The details required must be identified and discussed during the
System Management Review.

C.5.2 Application software packages

Each application package must be handed over with the appropriate
Hand-over Note, see section C.4.1. If an error is identified in the
software supplied, then a fix will be needed that identifies the error,
and the PinICL number or numbers involved.

C.5.3 Reference Data packages

Each application must be handed over with test versions of any
application specific Reference Data to allow the application to be
integrated into the Horizon system and the total system tested. If an

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-39 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

error is identified in the data supplied, then a fix will be needed that
identifies the error and the PinICL number or numbers involved.

C.5.4 Fixes

For any error to be resolved, information must be provided that allows
the validity of the fix to be established. The information required is
summarised in the following table:

tem \Description

Files to be delivered Identify the files, whether new or
replacement, extent of change.

Installation Instructions __IInstallation details including any process
dependencies.

(Additional testing needed (Definition of the regression tests needed to
prove that the fix is successful.

‘Known Problems lldentify any known issues.

Regression capability lldentify how the fix can be removed after it
lhas been applied.

Dependencies Desktop state (up or down) for a counter
application.

Batch files to be run For each file the following is required:

ie Name of batch file.

le Failure conditions and how they are
recognised.

le What regression capability is included
and how it is actioned.

le What rollback action is included.
le What tidy up action is included.
le Any testing instructions.

le Any tools needed.

Table C-7 Fix information

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-40 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.6 System Test

It is at this point in the life cycle that the application is tested in an end
to end system, with the same configuration as in the live estate.
Software packaging is an important element as this is the first point at
which the live configuration is tested. Security is also an important
element, because it may have operational implications. The way the
system is used in the outlet, as well as the support capability of the
application and its performance against predictions, are also tested at
this point.

Errors may be raised on any of these items. The deliverables for this
stage are therefore:

e Test scripts needed to prove that the functionality in an end-to-end
context.

e Operating instructions for users at the counter.

e Training course and Training Mode material that can be evaluated
against the application software.

C.6.1 Test scripts

These are needed to prove that the functionality provided by the
application functions as predicted in the Horizon environment. Test
Data Scripts must be provided that demonstrate the acceptability of the
software. There must be a series of scripts to prove that:

e Allinterfaces support the data predicted in the interface definitions
in the Application System Description.

e All operational elements of the application show the characteristics
identified in the Application System Description and the Application
Support Documentation.

e All events, error handling and diagnostic capability identified in the
Application Support Documentation are provided and operate as
predicted.

e Application interface and behaviour as presented to the clerk reacts
in the way predicted in the User Operating Instructions.

e All dependencies function as predicted in the Application
Dependencies Definition.

e Security components function as predicted by the Security Model.

e Application provides the performance characteristics predicted by
the Performance Model.

e Training Mode for the application provides the facilities defined in
the Application Training Documentation.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-41 of 45

FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.6.2 User Operating Procedures

These must cover the way the dialogue is conducted with the user in
normal, error, and contingency modes.

The way the dialogue is conducted with the clerk has implications for
the Counter Procedures and must be discussed and evaluated during
a Process Review conducted during the System Test activity.

C.6.3 Performance monitoring

Checks are made to ensure that the application performs as expected
from the Performance Model, and that when resources are monitored
the application is found to have the predicted impact on the throughput
and infrastructure performance of existing applications. This includes
the monitoring of resources across the live estate.

During System Test, the performance of the application against the
predictions in the Performance Model will be established. ICL Pathway
is responsible for the capacity management and performance
monitoring of the system, including any third party applications that
use the Horizon infrastructure. Consequently, the Performance Model
must define all the characteristics that support capacity management
and performance monitoring.

The live service is monitored for the gross behaviour of the shared
components, such as CPU usage, memory usage, and disc input and
output.

C.6.4 Training course

Any training course and Training Mode material developed will be
evaluated against the software delivered and issues raised resolved
via the normal error correction process using PinICLs.

C.6.5 Fixes

For any error that to be resolved, information must be provided that
allows the validity of the fix to be established, see section C.5.4 for
details.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-42 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.7 Application Installed and Implemented

This stage covers the installation throughout the live estate and the
initiation of its use by staff at each outlet. The co-ordination of the
installation of the software and Reference Data components, and the
user procedures and training schedule, are the key elements at this
point as is the ability to regress the implementation should errors
occur.

The main deliverables at this point are:

e Counter Procedures, based on the User Operating Procedures, that
include agreed manual as well as application procedures.

e Training plan that identifies how users will be trained in the user of
the facilities provided by the application.

Any issues identified in these areas must be addressed at a System
Management Review.

C.7.1 Counter Procedures

The current set of Counter Procedures must be updated to include the
facilities provided by a new application. These procedures include all
manual and system based procedures and identify when additional
support, such as calls to the Help Desk, should be invoked.

The Help Desk and support procedures must also be updated to match
the advice given in the Counter Procedures.

C.7.2 Training plan

This must cover all training activities for all the users of the application,
including how any Help Desk staff will be trained to answer the calls
identified in the Counter Procedures.

C.7.3 Fixes

For any error that to be resolved, information must be provided that
allows the validity of the fix to be established, see section C.5.4 for
details.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-43 of 45
FUJ00001377
FUJ00001377

ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000

C.8 System Supported

After the application has been introduced into the live estate, the
support of the application becomes significant as does its actual
performance and the way any error fixes are introduced into the estate.
The impact of the training the users receive becomes apparent in the
number of calls to the Help Desk.

The key elements at this point are:

e The support needed to resolve issues identified with application
functionality.

e Training received by the users and their use of Training Mode.
e Performance of the application when used by users at the outlet.

e How easily diagnostic information can be obtained and issues
resolved.

e How easily upgrades can be introduced when new functionality is
introduced.

C.8.1 Application support

In a distributed estate, it is important that diagnostic information can be
recorded by applications in files held on the counter PC that can be
interrogated centrally.

The level and kind of diagnostics depends on the application involved
and must be agreed with ICL Pathway during the System Management
Review prior to hand-over of any work packages as it can significantly
affect the support resource requirement.

C.8.1.1 Resource monitoring

Applications should not assume that there is any external resource
monitoring in place.

If resources require monitoring, the application must provide a self-
administering routine that takes appropriate action should the resource
exceeds its parameters and this must be raised at the Performance
Review.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-44 of 45
FUJ00001377

FUJ00001377
ICL Pathway Generalised API for OPS/TMS Ref: TD/STD/004
Appendix C System Management Version: 1.0
COMMERCIAL IN CONFIDENCE Date: 5/5/2000
C.8.1.2 Version numbers

It is recommended that applications define version information for each
DLL in the standard NT properties field belonging to the DLL.
Furthermore, it is recommended that when an end user application is
first loaded, it writes version information to the NT event log, and also
to an application specific diagnostic file.

C.8.2 Training

During live operation the training identified in the Training Plan is
delivered. The effectiveness of the use of training Mode will be
reflected in Help Desk calls. The effectiveness of both kinds of training
is continuously monitored by ICL Pathway, in terms of the impact on
Help Desk calls.

C.8.3 Performance monitoring

During System Test the performance of the application will have been
tested against the predictions in the Performance Model. In live
running the performance will be monitored.

C.8.4 Upgrades

For any upgrade that is to be delivered, information must be provided
that allows the validity of the upgrade to be established in the same
way as fixes are handled, see section C.5.4 for details.

C.8.5 Fixes

For any fix to be delivered, information must be provided that allows
the validity of the fix to be established, see section C.5.4 for details.

© 2000 ICL Pathway Limited COMMERCIAL IN CONFIDENCE Page: C-45 of 45