FUJ00123758 - Fujitsu Audit Data Retrieval High Level Design by Alan Holmes

Evidence on official site

FUJITSU

FUJ00123758
FUJ00123758

Audit Data Retrieval High Level Design
COMMERCIAL IN CONFIDENCE

Document Title:

Document Type:

Release:

Abstract:

Document Status:

Author & Dept:

Audit Data Retrieval High Level Design

High Level Design

N/A

This document describes the Audit data extraction & filtering
facilities within HNG-X

APPROVED

Alan Holmes

Internal Distribution:

External Distribution:

Approval Authorities:

Name

‘Adam Spurgeon

Role Signature Date

Solution Design

Graham Allen

Development

Alan Holmes

Architecture

Note: I See RMGA HNG-X Reviewers/Approvers Role Matrix (PGM/DCM/ION/0001) for guidance.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029

Version: V2.0
Date: 11/06/2010
PageNo: 1 of 42.
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

0 Document Control

0.1 Table of Contents

DOCUMENT CONTROL

Table of Contents

Figures and Tables.
Document History
Review Details
Associated Documents (Internal
Abbreviations
Glossary.
Changes Expected ..
Accuracy ....

INTRODUCTION...

ie Io

SISISISISISISE
SSISISISIERIE

ic
Ko

I=

In

leo

DESIGN PRINCIPLES...........

in

REQUIREMENTS ..........00000

SYSTEM OVERVIEW

meal

RISISE
Fara

SYSTEM COMPONENTS...

Application Components.
Audit Track Retrieve!

4

2 Audit Track Extractoi
3 Audit Extractor Clien
4 ARQ Database
5 Extraction Tools
6

7

8

9

im I

Filtering & Viewing
PAN Hash Calculation

PAN Decryption.
Data Delivery.

6.2 Distributed Application Service:
6.3 Networking Services. 36
6.4 Platforms
7 SYSTEMS MANAGEMENT... +38
©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref. DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 20f 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

7.4 Monitoring.
7.2 Schedule Req ments

8 APPLICATION DEVELOPMENT

9 SYSTEM QUALITIES
9.1 Availability ...

9.2 Usability
9.3 Performance

9.4 Security
9.5 Potential for Change ..

10 MIGRATION.

10.4 Audit Workst
10.2 Audit Servers

0.2 Figures and Tables

Figure 1- Audit system context
Figure 2- Interfaces to the Audit Track Retriever .
Figure 3 - Interfaces to the Audit Track Extractor
Table 1 - Requesters Data .
Table 2 - ARQ Data...
Table 3 - Horizon FAD to cluster mapping .
Figure 4 - Branch Transaction Query Overview .
Figure 5 - Branch Transaction Query Processor.
Figure 6 — Audit Distributed Application Interfaces
Figure 7 — Network Services

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 3 0f 42
FUJ00123758
FUJ00123758

co Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

i
OFFICE

0.3 Document History

Version No. Date Summary of Changes and Reason for Issue Associated Change -
CP/PEAK/PPRR
Reference
0.4 13/12/2006 Initial draft. Content taken from the equivalent Horizon
document SD/HLD/002
02 11/05/2007 Second draft. Includes changes resulting from comments
received on version 0.1 and PCI changes CP4305
03 02/02/2009 Changes resulting from review of version 0.2 tines Seooes
Changes resulting from E2E group review of the parent HNG-X CP0258
architecture document ARC/SVS/ARC/0001 HNG-X CP0284
Changes resulting from the following CPs
« HNG-X CP0157 — CP4623
* HNG-X CP0299 —- CP4810
+ HNG-X CP0258 ~ CP475t
* HNG-X CP0284 - CP4791
Extended migration coverage
04 44/10/2009 Clarification of Anti-Virus product used on the Audit PC0184551
workstation. PC0184537
PC0188605
Address comments received from review of V0.3
Addresses various Peaks (listed)
1.0 30/10/2009 Updated from comments received
Issued for approval
14 25/03/2010 Additional facilities within the Audit extraction system to enable I HNG-X CP0336
counter system events to be analysed and included as one of
the outputs of an ARQ
2.0 41/06/2010 Updated following review.
Issued for Approval

0.4 Review Details

NIA

Review Comments by

Review Comments to Alan Holmes & RMGADocumentManagement

Mandatory Review

Role Name

Solution Design

‘Andy Williams

Development Gerald Barnes
Development ‘Andrew Mansfield
Architecture Alan Holmes

Solution Design

David Harrison

ssc

Steve Parker

Business Continuity

Adam Parker

©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010
PageNo: 4 of 42
FUJ00123758
FUJ00123758

co Audit Data Retrieval High Level Design on
COMMERCIAL IN CONFIDENCE

SV&l Manager Chris Maving
LsT John Rogers
Role Name
Chief Information Security Officer Tom Lillywhite
Security & Risk Team CSPOA Securit
Programme Manager Geoff Butts
Applications Architecture David Johns
System Qualities Architecture Dave Chapman
Architect Jason Clark
Security Architect Tom Lillywhite
Test Design George Zolkiewka
Head of Service Management Gaetan van Achte
Head of Service Change & Transition Graham Welsh
Service Support Kirsty Gallacher
Service Network lan Mills
Data Centre Migration Vince Cochrane
Integration Team Manager Peter Okely
Testing Manager Debbie Richardson
LST Manager Sheila Bamber
RV Manager James Brett (POL, JTT)
POL Design Authority lan Trundell (POL, via Document Control)
VI & TE Manager Mark Ascott
Integrity Testing Michael Welch
Core Services Ed Ashford
Core Services ‘Andrew Gibson
Business Architect Gareth Jenkins
Development Graham Allen
Solution Design Architect Sarah Selwyn
Software & Solution Design Developer Stuart Honey
Service Manager - Retail and RMGA Claire Drake

Issued for Information — Please restrict this

distribution list to a minimum.

Position/Role Name

(*) = Reviewers that retumed comments

0.5 Associated Documents (Internal & External)

Reference Version Date Source
PGM/DCM/TEM/0001 Fujitsu Services RMGA HNG-X Dimensions
Document Template
©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref. DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 5 of 42
FUJITSU

Audit Data Retrieval High Level Design
COMMERCIAL IN CONFIDENCE

FUJ00123758
FUJ00123758

Reference Version Date Tith e
(DO NOT REMOVE)
SD/HLD/002 Audit Data retrieval HLD (Horizon) Pvcs
AD/DES/042 High Level Design of Common Pvcs
Agents
ARC/APP/RTM/0005, Requirements Traceability Matrix for Dimensions
Support Services
ARC/GEN/REP/0001 HNG-X Glossary Dimensions
ARC/SEC/ARC/0003. HNG-X Technical Security Dimensions
Architecture
ARC/SVS/ARC/0001 HNG-X Architecture — Support Dimensions
Services
CR/FSP/006 Audit Trail Functional Specification PVCS
DES/APP/HLD/0020 Branch Database High Level Design Dimensions
DES/APP/HLD/0030 Audit Data Collection & Storage High Dimensions
Level Design
HNG-X Audit Server (ARC) - Physical
DES/PPS/PPD/0035 Platform Design Dimensions
HNG-X Audit Workstation (AUW) -
DES/PPS/PPD/0037 Physical Platform Design Dimensions
DES/SEC/HLD/0002 HNG-X Crypto Services High Level Dimensions
Design
DES/SEC/HLD/0011 HNG-X Anti Virus High Level Design Dimensions
DES/SEC/IFS/0001 HNG-X Cryptographic Applications Dimensions
Programming Interface Specification
DEV/APP/LLD/0071 Audit Data Retrieval Low Level Dimensions
Design
DEV/APP/SPG/0016 Audit Extraction Client Support Guide I Dimensions
DEV/APP/SPG/0020 Audit Server Support Guide Dimensions
DEV/GEN/MAN/0015 Audit Extraction Client User Manual Dimensions
DEV/INF/ION/0001 Audit Server Configuration Dimensions
DEV/INF/ION/0008- Audit Server Migration Configurations I Dimensions
DEV/INF/ION/0009. Audit Server Schedule Design Dimensions
IA/PRO/004 Audit Data Extraction Process Pvcs
SD/IFS/014 Audit to Tivoli Cluster Information PVCS
Interface Specification
SVM/SDM/SD/0017 Security Management Service - Dimensions
Service Description
Extensible Markup Language (XML) ~ H
1.0 IRRELEVANT!
©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref. DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 6 of 42
FUJ00123758
FUJ00123758

Audit Data Retrieval High Level Design

oO
FUJITSU COMMERCIAL IN CONFIDENCE

Unless a specific version is referred to above, reference should be made to the current approved
versions of the documents.

0.6 Abbreviations

Abbreviation Definiti

ARQ Audit Record Query

AS Audit Server

ATE Audit Track Extractor

ATR Audit Track Retriever

ATS. Audit Track Sealer

BRDB Branch Database

COTS. Commercial off the Shelf

csv Comma Separated Values

EMC Company that supplies resilient disk technolog

FAD Finance Accounts Division

FS Fujitsu Services

FTMS File Transfer Managed Service

HNG Horizon Next Generation replacing current Baseline Horizon solution

HNG-X Horizon Next Generation — Plan X

KEL Known Errors Log

NBS Network Banking Service

NBX Network Banking Service — Horizon Implementation

NPS Network Persistent Store

NSP Atalla Network Security Processor

NBSC. National Business Support Centre

OBC Operational Business Change

PAN Personal Account Number

PCI Payment Card industry.

PCI DSS. Payment Card Industries Data Security Standard. A set of security controls defined
by the Payment Card Industry organisation.

PO Post Office

Post Office Post Office Limited

RAG Riposte Attribute Grammar

RDDS Reference Data Distribution System

RDMC Reference Data Management Centre

RDS Post Office Reference Data System.

RDT Reference Data Team - the Post Office and Fujitsu Customer Services teams use the
RDT environment to validate and verify the reference data associated with business
changes.

SLT Service Level Target

SYSMAN The systems management environment.

TES Transaction Enquiry Service

TMS Transaction Management Service. Used within the Audit system to refer to Riposte
audit data.

TwsS Tivoli Workload Scheduler

XML Extensible Markup Language

0.7 Glossary
See HNG-X Glossary (ARC/GEN/REP/0001)
m niti
Baseline Horizon the existing solution being re-architected
Branch Post Office outlet identified by a unique Branch Code. Within the HNG model, a
©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref. DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 7 of 42
FUJ00123758
FUJ00123758

co Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

OFFICE

Term Definition

Branch is a logical entity that can be composed of several physical locations at which
business is transacted

Cardholder Data The PAN, Cardholder name, Service code & Expiration date information relating to a
payment card

Hydra The name given to the live Post Office Account system during that period of the
HNG-X migration after the Data Centre Migration, but before the Branch Migration
completes

Operational Services Those services that are needed to run the Horizon system that are not directly

supporting the Post Office business. Examples include software distribution, audit,
security management etc.
Sensitive Authentication I The full contents of any track from the magnetic stripe (on the back of a card, in a

Data chip, etc.),
Any data element extracted from the magnetic stripe other than Cardholder Name,

PAN, expiry date and service code, and
Encrypted PIN blocks.

0.8 Changes Expected

Any future revision of this document should include an additional section which details the test impact of the
documented changes.

0.9 Accuracy

Fujitsu Services endeavours to ensure that the information contained in this document is correct but, whilst every
effort is made to ensure the accuracy of such information, it accepts no liability for any loss (however caused)
sustained as a result of any error or omission in the same.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 8 of 42.
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

1. Introduction

Within the HNG-X system, Fujitsu Services are required to provide facilities to produce, store and present
to authorised POL staff, Audit Track data in support of the security policy and audit requirements laid
down for the system.

The architecture for the audit sub-system within the HNG-X system is described in HNG-X Architecture -—
Support Services (ARC/SVS/ARC/0001). This Audit Data Retrieval High Level Design Specification is
consistent with that architecture.

This High Level Design (HLD) specifies the components required to provide the Audit Data Retrieval
facilities together with their interfaces and functionality. The level of detail in this HLD is intended to be
adequate to enable detailed design, implementation, integration and test work packages to be specified.

These Audit Data Extraction and Filtering facilities are provided for authorised Fujitsu Services staff to
provide extracts of the audit data from the Audit Archive in response to information requests from
authorised POL staff.

The Audit Data Extraction facilities are responsible for providing the facilities to filter, the retrieved audit
tracks to the level of detail specified in the information request.

The extraction of Audit Archive data is carried out by Fujitsu Services. The facilities that are provided for
the extraction of data that Fujitsu Services is contractually obliged to archive for audit purposes, will also
be available to extract non-contractual audit data.

This document also includes details of the changes to the Audit system required for CP4305. The only
impact of this change to the Audit system is that PANs, which are classified within PC] DSS as
‘cardholder Data’, are no longer stored in clear text form within Audit Tracks. Thus the Audit system
requires a mechanism to handle hashed and/or encrypted PANs where they form a part of retrieved Audit
data.

This version of the document includes details of changes for HNG-X CP0336. As part of the data
retrieval process, the archived events generated by counters at the branch are also analysed to identify
any possible occurrences of problems which might adversely affect the integrity of the transaction data.
HNG-X CP0336 integrates this process into the mainstream Audit retrieval applications.

This document is based upon the equivalent Horizon document SD/HLD/002.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 9 of 42.
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

2 Scope

The approach is to utilise where possible software products, already adopted or supplied with the
operating systems, to access, retrieve, filter and present audit data to the requestor. Extraction is to be
effected by the use of filters applied at various points in the extraction process.

This High Level Design Specification covers:
* Online extraction of Audit data from the Audit archive
e Seal Checking to ensure extracted files have seals intact
« Maintenance & Monitoring of Audit Record Queries (ARQs)
« Server based tools to filter high volume data
e Presentation to Auditor, tools to present data in required format

e The interfaces between the Audit server based applications restoring the Audit Tracks and the
Audit Data Extraction facilities on the Audit workstation

The data that will be stored in the Audit Archive and hence the data that needs to be retrieved is defined
in Audit Server Configuration (DEV/INF/ION/0001).

The scope of this HLD does not cover:

Specification of Information Requests, this is defined in Audit Extraction Client User Manual
(DEV/GEN/MAN/0015)

e Online access to live data to support Internal Audit
e The analysis/interpretation of Audit Tracks to provide specific Audit Trails

Minimum changes are expected to be made to the Horizon Audit applications as part of the migration to
HNG-X. Specific changes which are required for HNG-X and which are covered in this document are:

e Replacement facility to analyze Horizon Riposte Audit tracks
e New facility to analyze HNG-X Branch database Audit tracks
« Changes to handle hashed and encrypted PANs as part of CP4305

e Changes to support server side filtration of Tivoli event data which is integrated in with HNG-X &
Horizon branch data retrieval requests. HNG-X CP0336

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 10 of 42
FUJ00123758
FUJ00123758

Audit Data Retrieval High Level Design

oO
FUJITSU COMMERCIAL IN CONFIDENCE

3 Design Principles

The main principle of this design is to provide the required audit data extraction and filtering facilities
while minimising the impact on Applications within the HNG-X Subsystems. This HLD includes design
features to interface to and support existing system features.

The Audit Architecture as defined in HNG-X Architecture - Support Services (ARC/SVS/ARC/0001)
identifies the need to be able to cope with change as the usage of the HNG-X system develops,
especially as new applications and services are introduced. Thus a significant design principle is for the
Audit Data Extraction and Filtering system to be able to support the introduction of such new facilities with
minimum impact.

The extraction mechanisms must provide only that data that is appropriate to the role of the requestor.

To provide facilities which can be built upon and developed to provide enhanced audit facilities for future
releases.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 11 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

4 Requirements

Incoming requirements will be captured in the Support Services Topic Architecture SRS. The
Requirements Traceability Matrix for Support Services (ARC/APP/RTM/0005) will identify where the
requirement is explicitly addressed within an HLD.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 12 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

5 System Overview

HNG-X Architecture — Support Services (ARC/SVS/ARC/0001) describes the overall audit system design
for HNG-X. This section provides an overview of the design of the parts of the Audit subsystem
supporting the Audit Data Extraction functions. The function of the Audit Extraction system is to:

e Effect extraction of Audit archive data via an Audit Workstation. The Audit Workstation is
dedicated to this task.

e Extract the appropriate records using specified extraction criteria to give a manageable number
of files/records. Extracting a subset of records from an audit archive file is not provided for all file
formats.

e Provide facilities to manipulate and subsequently browse the extracted data to meet the criteria
of the ‘Information Request’

¢ Seal check retrieved files
e Maintain & Monitor Retrievals by ARQ

The source of the data to be extracted and filtered is generated from a wide range of components of the
HNG-X system including:

e Post Office Counter messages and transactions
e Systems Management Facilities — primarily application, system & security events
e Database Hosts (including the Reference Data System)
e FTMS Gateways
e System Scheduler logs
e Logging of activities undertaken by support staff while maintaining the system
e NPS database audit data
Access to the retrieval and extraction facilities is via the User Interface provided on the Audit Workstation.

Figure 1 gives an illustration of the major components of Audit sub-system on a single Campus. The
configuration is duplicated on both the live and standby data centres.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 13 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

Audit
Workstations :

(at BRA! & Atalla NSP- :
LEWo2) Audit Track
I -¥I Replication

Audit Server
Remote Link
ToDeta cente
Tivoli Event /
Management
System A.
oan 4
Secure
d Campus LAN ) Long
\ Term
. Storage
Other Audit i an.

Data Generating
‘Subsystems

Branch Host Database External System
Database Systems Gateways

Figure 1- Audit system context

5.1 Audit Server

The Gathering & Storage components of the Audit Server are defined in Audit Data Collection & Storage
High Level Design (DES/APP/HLD/0030).

The Audit Track Retriever component operates on the Audit server & provides support for the retrieval of
data from the Audit archive when requested by the Audit extraction components.

The Audit Track Extractor component, consists of a number of utilities that extract and filter data retrieved
by the Audit Track Retriever component

In overview the Extraction & Filter components of the Audit Server are:
e A utility to restore audit tracks from the Audit archive

e A utility to interrogate data, which originated from Post Office counters. Transaction records
relating to a specific outlet can be interrogated using a comprehensive query system.
Additionally the system events logged by counters at the outlet can be analysed to determine if
any problems occurred at the outlet which could cast doubt over the integrity of the transaction
data.

e A utility to interrogate data which relates to Network Banking transactions. Transaction records
which relate to a specific PAN can be isolated.

« A utility to manage ARQs as they progress through the retrieval & extraction processes. The
data to support this facility is maintained on the Audit Server.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 14 of 42
FUJ00123758
FUJ00123758

Audit Data Retrieval High Level Design

oO
FUJITSU COMMERCIAL IN CONFIDENCE

« Tools to assist in the analysis of hashed and encrypted PANs which have been stored within
Audit Tracks

e COTS tools to view and analyze ad hoc types of extracted data

« Tools to filter audited system events which originated from a given HNG-X or Horizon outlet plus,
in the case of HNG-X, the Branch Access Layer servers.

5.2 Audit Workstation

The Audit Workstation provides facilities for authorised Fujitsu Services staff to access the Audit Server in
order to retrieve Audit Track data from the Audit Archive and to either select or prepare Audit Track data
for presentation to POL or in support of internal audit activities. The Audit workstation is dedicated to this
task.

Access is via the standard user interface provided by Windows XP. A user Interface is provided to give
the user access to the applications on the Audit Server. A Retrieval User Interface is provided to search
for specific audit tracks & manage details of ARQs.

The files retrieved from the Audit Archive are stored on the Audit Server. Where necessary audit tracks
will be restored and filtered before files containing the relevant records are transferred to the Audit
Workstation. Additional browse and filter tools will be configured on the Audit Workstation where
subsequent searches/filters on files may be performed.

WordPad is provided to browse simple unstructured text files. Microsoft Office 2003 is provided to
analyze other types of file. In particular Microsoft Excel may be used to perform local analysis of
retrieved data.

User interfaces are provided on the Audit Workstation to the restore utilities. A detailed definition of the
utilities provided is contained in section 6 below.

There will be no automated synchronisation between the Audit Data Extraction and the Audit Data
filtering facilities.

The Audit Workstation supports a Write-Once CD to which selected Audit Track data for POL can be
written.

POL staff will not be given direct access to the Audit Workstation to safeguard other parts of the HNG-X
system. Instead nominated Fujitsu Services personnel will supply audit information as requested by Post
Office.

Audit workstations are sited at two Fujitsu sites — there are five Audit workstations at BRAO1 and two at
LEW02. Each of the two Audit workstation locations has a dedicated Atalla NSP which is used, by the
workstation, to decrypt PANs contained within Audit Tracks which relate to banking or payment card
transactions.

5.3 Audit Archive

The Audit Archive is held on an EMC Centera, this data is replicated to the remote data centre. Audit
Tracks generated on each campus are added to the local copy of the archive.

A more detailed description is defined in Audit Data Collection & Storage High Level Design
(DES/APP/HLD/0030).

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 15 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

5.4 Audit Points

An Audit Point is a logical concept introduced in this design to minimise the linkage (as seen by users of
the Audit Workstation) with the physical design of the HNG-X system. This is intended to help reduce the
knowledge that auditors need of the details of the way HNG-X has been implemented.

The term Audit Point is used, in a number of places within this design, to refer to the logical location at
which a particular Audit Track is generated. Due to the distributed and resilient nature of the HNG-X
system an Audit Point is actually realised at a number of different physical locations.

The specific locations at which the Audit Track of a particular Audit Point is generated are identified as
Audit Sub-Points. An Audit Sub-Point maps onto a single sub-directory on a single component in the
HNG-X system. It is however possible for an Audit Sub-Point to map onto a (finite) set of such sub-
directories. Where there are a number of sub-directories they will be nested beneath a single top-level
sub-directory.

The files stored in the Audit Archive are named in terms of an Audit Point and an Audit Sub-Point. These
logical concepts are designed to be stable across the life of the HNG-X system and to assist the Fujitsu
Services Auditors in locating the files containing the relevant data to support any particular audit activity.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 16 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

6 System Components

6.1 Application Components

Audit Data Retrieval is handled by the Audit Track Retriever and Audit Track Extractor components of the
Audit solution. These are described in the following sub-sections.

6.1.1. Audit Track Retriever

The Audit Track Retrieved is a server side bespoke application which supports the retrieval of Audit
Tracks from the Audit archive.

Figure 2 identifies the main interfaces to the Audit Track Retriever (ATR)

Tivoli M it LATR-L taps.
ivoli Managemen -ATR -ATD-
Envionent I AIDS pe Audit Track Deleter
Audit Track Retriever
L-ATR-2/
I-ATS- I-ATR-4
Audit Track Sealer <—LATS6_ <4 Sealer Database
LATR-6/
L-ATR-7 I-ATE-2

Fujitsu Services Auditors Audit Track Extractor

Figure 2- Interfaces to the Audit Track Retriever

6.1.1.1 ATR Interfaces

I-ATR-1 Interfaces to the system administration staff and facilities are all via the Tivoli Management
Environment (TME). TME shall monitor the existence of the ATR and shall report as a Tivoli event any
unexpected unavailability.

The ATR will be under control of TWS scheduling.
In addition command line interfaces will be provided to start and stop the ATR outside of TWS control.

I-ATR-2 Files whose seals are to be checked are notified to the ATS via this interface. Each request is in
the form of a marker file that uniquely identifies the filename of the file to be seal checked. Files to be
seal checked are placed in an agreed directory. The ATS regularly checks the directory and removes the
first entry and checks the seal. The ATS is responsible for removing files from the common directory.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 17 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

1-ATR-3 For every file successfully (or unsuccessfully) retrieved from Centera by the ATR it will inform the
Audit Track Deleter (ATD) of:

* The time and date of the Retrieval
e Ameaningful success/failure code
e The path name file

The ATR will pass the details of the files retrieved to the ATD via an agreed directory. Details of the files
will be put in a Record File in the shared directory for collection by the ATD. Each Record File must
contain the details of at least one file, but may contain many. Details of retrieved files batched together in
one Record File must always be held on persistent storage media (e.g. disk) and must never be over
written. Record Files must regularly (e.g. every hour) be passed over to the ATD. The frequency shall be
a configurable parameter. The Record File shall be a text file.

Note that this interface is to support internal logging of Audit Server activities.

I-ATR-4 This interface is to the Sealer SQL Server 2000 database. This database contains details of all
archived Audit tracks held by the Audit server.

1-ATR-6 The files retrieved from the Audit Archive are placed into a directory from which the Audit Track
extractor can copy the data for subsequent extraction activities. The ATE is responsible for deleting the
files from this directory. This deletion will occur when the relevant ARQ is closed.

I-ATR-7 represents the ‘Retrieval User Interface’, which Fujitsu Services auditors use to retrieve files
from the Audit Archive. This interface is one part of the graphical user interface provided for Data
Extraction

6.1.2. Audit Track Extractor

The Audit Track Extractor (ATE) component of the Audit solution will be implemented as a bespoke set of
functions available from the Audit workstation to refine & interrogate data retrieved by the Audit Track
Retriever (ATR).

The level of extraction will always be dependent on the associated Audit Record Query and the
characteristics of the audit archive file containing the data being requested, e.g. an external interface
Control File which has been archived in its original format, is likely to be requested in its entirety with no
further extraction of specific records required. Files containing transaction data generated by Post Office
Counters, due to how they are generated, may contain data associated with many outlets, are less likely
to be required in their entirety.

The different levels of filters to achieve extraction of required data are
e File Identification, by Audit Point, using data from ARQ and Audit Server Configuration

e Extraction of a subset of audit tracks from a set of audit points file using defined date based
extraction criteria

e Filtering (post extraction or post retrieval if no further extraction took place) using tools
appropriate to the format of the extracted file

The functionality provided will comprise of: -
« Management and Monitoring of Audit Record Queries (ARQs)

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 18 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

¢ Online extraction of data from the Audit Archive

e Seal Checking to ensure extracted files have seals intact

* Server based tools to filter Audit Tracks

e Workstation based tools to analyze and present audit data in required format

e Ability to copy extracted data to CD-ROM for delivery

6.1.2.1 ATE Interfaces
Figure 3 identifies the main interfaces to the Audit Track Extractor (ATE)

Tivoli Management LATE-1 LATE-7
+ _EATET yop.
Environment - CD-ROM Discs
Audit Track Extractor
L-ATE-3 T-ATE-2/
RFi Database = <> ag :ATR6___ Audit Track Retriever
T-ATE-4
L-ATE-6
Seal Database
Extraction
Directory

Figure 3 - Interfaces to the Audit Track Extractor

I-ATE-1 Interfaces to the system administration staff and facilities are all via the Tivoli Management
Environment. This consists of remote access for administration purposes and event monitoring.

I-ATE-2 (I-ATR-6) the files retrieved from the Audit Archive are placed into a directory from which the
Audit Track extractor can copy the data for subsequent extraction activities. The ATE is responsible for
deleting the files from this directory.

I-ATE-3 This is the interface from the workstation applications to the SQL Server ARQ Database where
ARQ and associated file information is stored. Updates to ARQ details are made from the workstation
applications.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 19 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

1-ATE-4 This is the interface to link to the Audit Seal Database (This database contains details of all
archived Audit tracks held by the Audit server) to allow the status of Seal Checks to be interrogated via
the GUI.

I-ATE-6 This is the interface between the Auditor using the tools to View and Filter data held in the
extracted directory on the Audit Server. Filtered audit track data will pass across this interface.

I-ATE-7 This is the interface between the Auditor and the CD-Writer device to write analysed audit track
data for use by external Auditors. This interface is implemented using standard Windows XP CD burning
functionality.

6.1.3 Audit Extractor Client

The Audit Extractor Client application that will reside on the Audit workstations handles the selection,
retrieval and recording of the usage of Audit Data. Access to the Audit Extractor Client is provided by a
Start Menu option.

The User name associated with the extractor client session is taken from the current logged on user
name; there is no requirement for the user to log on to the extractor client.

The GUI will initially provide options to
« Create anew ARQ
e Open an existing ARQ
e Close an existing ARQ
e Monitor the progress of server based retrieval & analysis tasks

Details of ARQs are stored on a Microsoft SQL Server 2000 database resident on the Audit Server's local
storage.

6.1.3.1 Creation of aNew ARQ

When an ARQ is received it should be registered at the Data centre for which the extraction is to be
performed by using the create ARQ function.

The user will be required to enter various control information relating to the ARQ: -
e The data centre at which the retrievals are to be performed
e Requester Id, from a selection of pre-defined values comprising:
o Other 3rd Party
o POA Internal Audit

o POA Other

o POASSC

o POL Internal Audit
o POL Other

o POL Security
« Catalogue Entry - Optional

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

Page No: 20 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

e Receipt Reference — Requester reference
e Request Date - The date the request was received.
e Required Date — The date that the information is required.
e Access Reason — Explanatory text
On completion of the above fields the user will be able to: -
e Save the ARQ, which will result in generation of a new ARQ Id
« Specify details of the files required to satisfy the ARQ.

e In the case of a retrieval of branch transaction data, specify whether system events for the
branch are also to be retrieved & analyzed.

« Cancel the request

6.1.3.2 Creation of a New Fast ARQ

As an alternative to opening a new conventional ARQ, a new ARQ may be opened in ‘Fast mode’. This
offers a much simpler user interface & a more streamlined workflow where branch transaction data is
being retrieved.

When a new Fast ARQ is opened, a single form is presented allowing the following information to be
entered.

e Requester Id —- See above

e Receipt Reference —- See above
e Request Date - See above.

« Required Date - See above

Filter start & end dates - The data range for the analysis. For extraction purposes, there is an
option to add extra days onto the end date to allow for where data has been gathered late.

« FAD code — the Branch Accounting code
e Whether system events for the branch are also to be retrieved & analyzed.

The set of queries to be run against the retrieved data.

Output Folder - The location (on the audit workstation) where the resulting spreadsheets are to be
written

Once all required data is entered, the ARQ may be executed & will, in the background:
e Retrieve all the data required
e Run the requested queries
¢ Execute the event filters if selected
e Generate the required spreadsheets on the workstation.
Status information will be displayed on the user interface showing progress through the above phases.

If a Fast ARQ is closed before it has reached the end of its normal execution, it will not be possible to
subsequently reopen it and allow it to continue. It will be necessary to close the ARQ and open a new
one.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

Page No: 21 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

It is anticipated that the Fast ARQ approach will be usable for the majority of retrievals of branch
transaction data. There will however be exceptional cases where the finer control offered by the
conventional ARQ is required.

6.1.3.3 Opening an Existing ARQ

The user will be able to select existing ARQs (at a given data centre) and open them to progress file
retrieval and analysis.

A list of the ARQs available at the selected data centre will be presented to the user. This list will only
contain open ARQs

Once an ARQ is selected & reopened, the user is able to progress the ARQ through to completion.

Note that if an existing ARQ was originally opened as a Fast ARQ, if it is closed & reopened, no further
updates may be made to the ARQ. When it is reopened the Fast ARQ form is displayed, read only with
all originally entered data.

6.1.3.4 Closing and Cancelling an ARQ

The user will be able to select existing ARQs (at a given data centre) and close them, preventing any
further activity on the ARQ.

Alist of the ARQs available at the selected data centre will be presented to the user. This list will only
contain open ARQs.

No further amendment to an ARQ will be permitted after it is closed.

It is the responsibility of the User to ensure that they have completed extraction and filtering of retrieved
files before they close an ARQ.

When an ARQ is closed ALL of the data files retrieved for the request will be deleted from the shared
Userarea filestore on the Audit Server along with other internal files relating to the ARQ.

An audit file (which is written to the Audit archive as part of the Audit servers own internal audit trail) will
be produced giving the following details of the ARQ: -

¢ Basic details of the ARQ
« Alist of ALL files associated with the ARQ and their seal status

e A\list of All of the actions taken by the user to complete the ARQ

6.1.3.5 ARQ Status Monitoring

The ARQ status display will show current state of all open ARQs (across both data centres). The status
values shown will only relate to server based actions undertaken on behalf of each ARQ, this will include:

e Audit Track retrieval
e Audit Track seal checking
e Server based filtration & analysis

Only the status for the most recently initiated server action will be shown on the initial display. It will be
possible to drill down into an individual ARQ to show additional information relating to preceding server
actions.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 22 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

6.1.3.6 File selection criteria

The information on the ARQ is used to determine the required files to be retrieved from the Audit archive
store

The Audit Server Configuration (DEV/INF/ION/0001) specifies the Audit points and sub-points known to
the Audit system. This enables the user to translate the User specified data from the Information
Request into a set of file selection criteria. There is no automated support for the translation process.

The required Data is specified on the File Selection screen invoked for a New ARQ and when a request
to amend selection criteria is made for an existing ARQ

The following selection criteria for file selection are required:

« Asingle Date Range within which the required files were collected by the Audit system.
It should be noted that files and data are not always available to the Audit system for some days
after their creation date.

e Optionally, a selection of Audit Points and Audit Sub-Points.
This selection to be controlled by lookup lists of all the available points.
In the case of Audit Sub-Points the list is to be restricted to those related to the selected Audit
point.

e Optionally, a file name template may be supplied. This allows the user to restrict the search to file
names that match the pattern entered. Wild cards may be included in the supplied pattern using
the * character.

« Optionally, a branch FAD code. This will automatically determine the relevant Audit point / Sub-
Points which contain transaction data generated by the outlet. This will require retrieving a
subset of data from the audit points containing both Horizon Correspondence server data and
HNG-X data Branch database data depending upon the date the branch migrated to HNG-X.
See sections 6.1.3.6.1 & 6.1.3.6.2

e In the case of a retrieval of branch transaction data (i.e. where a FAD code was entered), specify
whether system events for the branch are also to be retrieved & filtered.

Once the above fields have been supplied the user may initiate the search for files that match the search
criteria.

6.1.3.6.1 Branch Cluster Determination — Horizon Correspondence Server Data

Because of the volumes of TMS files, it is a requirement to retrieve only those files relevant to the Outlet.
This necessitates being able to establish which Correspondence server cluster an outlet's messages will
have been created on for the requested Date Range.

For Horizon counters, this information will be provided by a Tivoli Web service. This returns historical
Cluster information for an Outlet. The interface to this view is defined in Audit to Tivoli Cluster Information
Interface Specification (SD/IFS/014).

e The Tivoli Web page will be accessed from the Audit Server

e The outlet required will be provided as a six character branch Id Code (i.e. without the trailing
check digit)

e The utility must provide ALL of the Correspondence Server Clusters in which the Outlet has
resided together with the corresponding date ranges

From the Tivoli provided information the Cluster Determinant component will determine the required Audit
Sub-Points for the requested date range.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

Page No: 23 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

The Audit system will require continued access to this branch to cluster mapping information for seven
years after the last Horizon counter is migrated to HNG-X. The Tivoli cluster information server used by
the Audit system will disappear at the end of Hydra. Until this time, the cluster mapping data will be
volatile and the Audit system will require access to the dynamic data held within the Tivoli cluster
information service. After this time, the data will be static and it will be imported into the Audit ARQ
database. The Audit system will then look up the mapping directly from its database. See section 6.1.4.6

Within each cluster, TMS audit data is further divided into eleven separate streams based on application
of a hash function to the Branch Id. There are four Correspondence server clusters, thus the set of files
that need to be retrieved to contain all data relating to a branch is =1/44" of the total volume of TMS audit
data for the period under analysis. The Branch ID hash function returns the following results:

«  'X'- for non outlet data
e [0-9] - for branch data based on the 3" digit of the 6 digit Branch Id

The hash value is embedded in the Audit Track file names generated by the harvesting agents. This is
further described in High Level Design of Common Agents (AD/DES/042)

This function is only applicable to data generated by Horizon Counters.

6.1.3.6.2 Branch Hash Value Determination - HNG-X Branch Database Data

An equivalent function to that described in section 6.1.3.6.1 is also necessary to limit the volume of
retrieved data which originated at HNG-X outlets.

Within the HNG-X Branch Database each branch is allocated a hash value ranging from 0 to 127. This is
used internally within the BRDB to partition data. The hash allocation remains constant for the life of the
branch. This hash value is also used to partition the audit data generated by the BRDB. A separate
audit stream is generated for each of the 128 possible hash values, each stream containing audit data for
all branches which map to the hash value associated with the stream. Data within a given stream may be
further partitioned, to ensure that an individual Audit Track file size is within the advisory limit for the Audit
system of =100MB. This process is further described in Branch Database High Level Design
(DES/APP/HLD/0020).

When the Audit retrieval system needs to retrieve BRDB data relating to a particular branch, it derives the
branch hash value from the branch Id & only retrieves (from Centera) data for the correct audit stream.

The audit system derives the branch hash value, from the branch Id, by direct look up from table
brdb_fad_hash_outlet_mapping in the BRDB.

This function is only applicable to data generated by HNG-X Counters.

6.1.3.7 File Retrieval facilities

The files that match the search criteria are displayed on the ARQ screen once a search has been
initiated.

This screen provides all of the functionality to manipulate the files associated with an ARQ.
The following facilities are to be provided: -
« Restore a selected set of files from the Audit archive to the Userarea of the Audit Server
e Check the status of a selected set of files (i.e. where they are in the retrieval cycle)
« Delete a selected set of files from the ARQ
e Replace the files for the ARQ with a selected set (i.e. delete all but)

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

Page No: 24 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

« Amend the search criteria which will remove the file list associated with the ARQ and present the
user with a file selection screen

e Add files which will effectively widen the search criteria and produce additional files on the list
e Allow the user to Monitor the progress of any actions being carried out for the ARQ
e Provide menu options for the server based processing of retrieved data

Once a file has been retrieved it needs to be automatically passed onto the ATS (for seal checking)

6.1.4 ARQ Database

This section contains a brief overview of the content of the ARQ Database. For full details of the
database structure sees Audit Data Retrieval Low Level Design (DEV/APP/LLD/007 1)

6.1.4.1 Overview
The ARQ database (aka RFI database) will be a SQL Server Database.

A separate database will be held on each Audit Server with data only pertaining to ARQ requests handled
by that server.

The database will be regularly backed up as part of the Audit Server backup schedule.
The ARQ database will contain the following information: -
e Details of information Requesters
¢ Details of Requests for Information (ARQs) and their associated files and query requests
* Details of valid users of the Audit facilities
e Details of Audit points and Audit Subpoints
e Horizon FAD to cluster mapping (post Hydra)

e Client-Server interface management

6.1.4.2 Requesters Data
The data recorded in the database for each requester will be: -

Attribute Description

Requester Identity Identifies the Requester of information.

Telephone No The telephone contact number of the requester

ARQ Prefix The prefix to be part of the unique number generated for each
ARQ. This will aid identification of ARQs by different requesters

Organisation The organisation to which the requester belongs (e.g. POA,
POL etc.)

Table 1 - Requesters Data

As it is expected that data in this table will be relatively static, maintenance will be via direct access to the
table using standard Microsoft SQL Server tools.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 25 of 42
FUJ00123758
FUJ00123758

Audit Data Retrieval High Level Design

oO
FUJITSU COMMERCIAL IN CONFIDENCE

6.1.4.3 ARQ Data
The Data recorded for each ARQ will be: -

Attribute Description

ARQ Reference A unique generated reference number for the ARQ constructed
from the ARQ prefix for the requestor and a sequential numeric
suffix

Operator Id The identity of the Auditor performing the ARQ taken from the
Logon information for the ARQ GUI

Requester ID The identity of the requester entered by the user from the
dropdown list presented to the user when creating an ARQ

Receipt Reference The requesters reference for the ARQ, input by the user

Access Reason Textual Details of the ARQ input by the User

Catalogue Entry Reference for the Auditor to relate to the manually held ARQ
catalogue

Date Received The date on which the ARQ was received, input by User

Date Required The date by which the information is required by the Requester,
input by the User

Status The status of the ARQ which is maintained automatically by the
system

Selection Criteria Details of the Audit Data selection criteria as input on the
Request for Information screen

Events filtration Whether Event data is to be included in the analysis

Fast ARQ flag Indicates that this ARQ is a ‘fast ARQ’

Table 2 - ARQ Data

6.1.4.4 Audit Points and Sub-Points

Tables of valid Audit Points and sub-Points will be held in the ARQ database to be utilised for the
provision of dropdown lists in the Retrieval User Interface.

These tables will be synchronized automatically (on a daily basis) with the Audit Server Configuration to
ensure that they contain an up to date view of the system configuration.

6.1.4.5 ARQ File and Query Status
The user will be able to monitor the status of files and queries utilised to satisfy an ARQ

6.1.4.6 Horizon FAD to cluster mapping

When performing Horizon TMS data retrievals, the Audit system will require access to branch to cluster
mapping information. The Tivoli cluster information server used by the Audit system will become obsolete
when all SYSMAN2 functionality is removed from the solution. Until this time the Audit system will
require access to the data held within the Tivoli cluster information service. After this time, the data will be

©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 26 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

static and it will be imported into the ARQ database. The Audit system will then look up the mapping
directly from its database.

The following data will need to be modelled in the database:

I Defin
FAD Six character FAD code of the branch.
Status The status of this branch (at the point of data import). Possible values are:

NOT_FOUND - FAD not found in Oracle table.

NOT_LIVE — No installation date (with no removal date) for any counters.
CLOSED - Set to permanently closed by OCMS.

LIVE — At least one counter is installed (with no removal date).
TEMPORARY_CLOSURE - Set to temporarily closed by OCMS.

Cluster Id Cluster id in range 1..4
Start Date The date the branch was assigned to this cluster
End Date The date the branch was assigned removed from this cluster. Not present if

this was the active cluster at the time of data import.

Table 3 - Horizon FAD to cluster mapping

See Audit to Tivoli Cluster Information Interface Specification (SD/IFS/014) and section 6.1.3.6.1 for
further information.

6.1.4.7 Client-Server Interface

Tables to manage the client-server interface between the Audit workstations and the Audit server will
support the following functionality:

« Recover files

e File status

* Get cluster/hash Id

e Query branch transaction data
e Query NBX data

* Close ARQ

« Create ARQ

6.1.5 Extraction Tools

6.1.5.1 Uncompressing Files

Files which have been compressed at source prior to being gathered by the Audit archive will require
manually decompressing prior to analysis or viewing. WinZip will be provided on the menu for the Audit
workstation for this activity.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

Page No: 27 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

An exception to the above rule is Branch database counter transaction data which is automatically
decompressed during retrieval by virtue of the PreCompress parameter in the Audit server configuration
file. See Archive Server Configuration DEV/INF/ION/0001

6.1.6 Filtering & Viewing

6.1.6.1. Branch Transaction Data Query

This feature enables branch transaction data for a specified Branch to be analyzed using customisable
queries. A set of branch transaction Audit tracks (which could contain Horizon and/or HNG-X branch
database data) must have been retrieved & seal checked prior to invoking this function.

« Manage sets of user queries

e Confirm the integrity & completeness of the data

e Execute equivalent queries against both Riposte & HNG-X transaction data
e Perform System event filtration

e Import the query results into Microsoft Excel workbooks.

All functionality that involves processing large volumes of retrieved audit data will be deployed on the
Audit server in order to minimise the volume of data that needs to be transferred to the Audit workstation.

The facility is described in greater detail in Audit Data Retrieval Low Level Design (DEV/APP/LLD/0071).

This is process is summarized in Figure 4

Audit Workstation Audit Server
(Nerul Process} Fitered
Events
Query
Results ~
Data Integrity
Report
ry

Query Submit Query
Management

“ties )
rocessor Fillers
—/

Riposte HNGX
User Queries Transaction Transaction Event Data
Data Data

Figure 4 - Branch Transaction Query Overview

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

Page No: 28 of 42
FUJ00123758

FUJ00123758

Audit Data Retrieval High Level Design

oO
FUJITSU COMMERCIAL IN CONFIDENCE

6.1.6.1.1 User Query Management

The Audit workstation will enable the user to maintain a set of queries that can be executed against
branch data. The user will be able to perform the following management operations on those queries.

e Create new queries

e Save named queries to filestore

« Load named queries from filestore

e Edit queries

« Maintain a set commonly used queries which can be quickly loaded via selection from a list

The user will be able to select the location where queries are to be stored to or loaded from. Queries will,
in general, be shared between all audit workstations - thus the user will typically store queries centrally.

Matched pairs of queries will be managed, one of the pair will operate on HNG-X transaction data and the
other will perform an equivalent query upon Riposte transaction data.

6.1.6.1.2 Query Submission

When, as part of an ARQ, a set of audit tracks have been successfully retrieved and seal checked, the
user may load a query and execute it against the audit tracks associated with the ARQ.

The query will be submitted to the Audit server based Query Processor. Execution of the query will
operate asynchronously. The user will monitor progress of the query using the standard Audit Extractor
mechanism used to check the status of server actions.

Queries will be automatically constrained to only return data for the single Branch Id that was used to
select audit tracks for retrieval — see section 6.1.3.6.

Once the server has completed the execution of the query, the user may open the results of the query
using Microsoft Excel. If required, any further analysis of the data may be undertaken using standard
Excel features.

The server will also generate a report on the integrity and completeness of the source data. This will be
included in the Excel spreadsheet as a separate worksheet.

6.1.6.1.3 Query Processor

The Audit server based Query Processor will accept and process Branch transaction data queries from
the Audit workstation. The process will support the following functionality:

« Operate within the context of an ARQ

e Accept a query and branch Id as input

e Process both HNG-X and Riposte Audit tracks, producing separate reports for each
« Perform an initial filtration of the audit tracks

e Perform Event filtration

« Perform integrity checks on both types of audit tracks, producing summaries of the results of the
check. See §6.1.6.1.3.3

e Apply the supplied queries to the filtered data, producing Excel compatible results files. Separate
results files are produce for HNG-X & Horizon data.

This logical view of this functionality is given in Figure 5

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 29 of 42
FUJ00123758
FUJ00123758

Audit Data Retrieval High Level Design

oO
FUJITSU COMMERCIAL IN CONFIDENCE

I Branch Transaction Query Processor

«Manual Process)
(Manual Process} ;

Horizon Query
Results

HNG-X Query
Results

XML Query Fitered HNG-X Fiered

‘3 Events Horizon Events

- - ry

pose Data
iment Report
GX bata
vy Integrity Report
ut

FING Data jposie ala (Riposte Data ‘Sysmand ‘Sysmiand ‘Sysman?
Integrity Integrity XML HNG-x Event Horizon Horizon
Check Check Conversion Filter Event Fitter ) \ event Filter

ry

TNGX Daa) —(Riposte Data
Tivabon Pivaton
I very Handee

-———_ ry ry
ARO
Database
pos FRGK— (5,
yemana I Sysmand
ma Transaction I Transaction I event ata I Event Data
Serve

Figure 5 - Branch Transaction Query Processor

6.1.6.1.3.1 Query Manager Service

The Query manager is a windows service that monitors the ARQ database for new query requests &
launches new Query handler instances to process them

6.1.6.1.3.2 Query Handler
The Query Handler manages the processing of the query. It is responsible for:
e Accepting the initial request
« Maintaining the status of the process in the ARQ database
e Differentiating Riposte & HNG-X audit tracks
« Manage the subsequent processing of the data

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

Page No: 30 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

As HNG-X and Riposte audit data will be fundamentally different in both structure and field naming, each
query will be composed of two parts — a Riposte and a HNG-X version. The correct part of the query will
need to be passed to the query engine according to the type of audit track being processed.

The end user will not need to be aware of the Horizon - HNG-X migration date of the branch in question.
The system will determine this by querying the Branch database and retrieve the appropriate type of
Audit tracks.

When retrieving data relating to branch transactions, the retrieval system must potentially retrieve both
Horizon & HNG-X Audit tracks for the branch & date range specified if the query dates span the branch
migration to HNG-X. The date that a Branch migrated from Horizon to HNG-X is available in the Branch
database.

6.1.6.1.3.3 Data Integrity Checks
The following integrity checks will be applied to the data

e Completeness of data — contiguous message sequence numbers
« — Integrity of individual messages
o For Riposte data the message CRC should be checked
o For HNG-X data the message signature will be verified

Separate Riposte & HNG-X summaries of the results of the integrity checks are generated. They should
detail:

« Summary of the message sequence runs broken down by counter Id. This should include start &
end date/times and start & end message sequence numbers. Any gaps in the message
sequence runs must be highlighted.

« Summary of messages that have failed individual message integrity checks

Any failure of the data integrity checks will not prevent subsequent execution of the query. The audit
workstation user will be warned of the failure via the server process status notification mechanism.

6.1.6.1.3.4 Riposte XML Conversion

Riposte audit tracks will be Riposte Attribute Grammar (RAG) format. In order that a common query
engine may be used, they will be converted to XML format. The following considerations need to be
made during this conversion.

Valid XML element names must conform to the following:

NameStartChar::= ":" 1 [A-Z] I "_" I [a-z] I [#xCO-#xD6] I [#xD8-#xF6] I [#xF8-
#x2FF] I [#x370-#x37D] I [#x37F-#x1FFF] I [#x200C-#x200D] I [#x2070-#x218F] I
(#x2C00-#x2FEF) I [#x3001-#xD7FF] I [#xF900-#xFDCF] I [#xFDFO-#xFFFD] I
[#x10000-#xEFFFF]

NameChar z:= NameStartChar van "." I [0-9] I #xB7 I [#x0300-#x036F] I

[#x203F-#x2040
Name ri= NameStartChar (NameChar) *

The rules for RAG attribute names are less restrictive - They can be comprised of any printable
characters excluding those in the set {whitespace I ‘:' I ‘.’ I ‘<' I >'}. Thus some RAG objects will need to
be renamed during conversion.

Valid data within an XML document consists of characters in the following range:

#x9 I #xA I #xD I [#x20-#xD7FF] I [#xE000-#xFFFD] I [#x10000-#x10FFFF]
©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref. DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 31 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

There are isolated examples within existing Riposte data where RAG attribute values contain non-
printable control characters outside of the above range. These cannot be represented within XML
content, but should be represented in a textual form (e.g. __xnn_;)

See W3C document Extensible Markup Language (XML) 1.0 f IRRELEVANT
detail on the rules for creating well formed XML documents.

for further

6.1.6.1.3.5 XML Query Engine
The XML Query Engine must provide at least the following functionality:

e — Identification of individual fields within a message via a path specifier
e Enable a subset of fields within a message to be selected for output
e Meaningful column headings can be included in the output

e Can handle mixed message types

« Awareness of the following basic data types

o String
o Numeric
o Date

« Output can be sorted on selected fields. The sort process should be type aware
e Supports the following expression types which are used to specify which messages are selected
o Existence of a field within a message

o Type aware comparisons using the following operators ({‘
Comparisons can be against fields or typed literal values

o Grouping of selection sub-expressions via parentheses to alter the order of evaluation

6.1.6.1.3.6 Event Filters

The Event filters are implemented as Perl scripts. Perl provides a sophisticated regular expression
implementation that is capable of filtering on complex criteria. A script based solution has been chosen
as it allows the filters to be more readily adapted to include or exclude particular events as required.

The scripts operate by filtering out all events originating from counters at the branch under investigation,
then excluding a set of known to be benign Horizon and benign HNG-X events. The resulting sets of
events are written to the output file in CSV format. These files can then be manually copied to the Audit
workstation and examined using Excel.

There are two sources of Tivoli event audit data, the Horizon OMDB (Sysman2) and the HNG-X EDS
(Sysman3) database servers. The format of the audit files produced by these two servers is very different
& each needs to be processed by its own filter script.

Note that although Horizon counter events will normally appear in the OMDB audit subpoint, once the
branch has had its branch router installed, events from the branch go up the Sysman 3 EDS route. Thus
when filtering events for a Horizon branch, both sources of audit data must be searched.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 32 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

6.1.6.2 View/ Filter Other Audit Track

The extraction of other (than Branch transaction) data will be via the filters, views, and search facilities
provided with COTS tools. Different filtering facilities need to be configured on the Audit Workstation, to
allow the filtering of the restored data to the level requested on the ARQ.

A text-based utility, MS Word's WordPad, will be used to view and search the retrieved files if required.

Windows Explorer also has the capability to search for data strings in many files

6.1.6.3 NBX Transaction Journal Filtering / Viewing

This feature enables high volume Network Banking transaction journals to be searched for entries related
to a user specified PAN.

With the introduction of CP4305 into the solution, the NBX journals will no longer contain the PAN in clear
text. Instead there will be an encrypted version of the PAN and, in a separate field, a securely hashed
version of the PAN. Historic NBX Audit Tracks which are held in the Audit archive will continue to hold
the PAN in clear text form.

To enable filtering of NBX transaction journals, they must first be retrieved to the Audit server using the
standard Audit Track Extractor. At the point they are extracted from the NPS, the NBX audit tracks are
partitioned using a hash function based on the PAN associated with the transaction - 64 separate
streams of audit data are written. The hash value is embedded in the Audit Track file name generated by
the NPS harvesting process. Thus, assuming a full clear text PAN is available, it is possible to only
retrieve the subset of Audit Tracks that contain transactions relating to the required PAN.

The hash function that is currently utilized on Horizon is:
(PAN MOD 64) +1
which yields results in the range 1-64.

With the introduction of CP4305, the full PAN is not available thus an alternative function is applied to the
hashed version of the PAN:

(RIGHT(PAN, 3) MOD 64) + 1

A given retrieval will need to evaluate both possible hash values & retrieve both the corresponding
streams of audit data.

Even allowing for this partitioning of data a 30-day enquiry will involve searching a large volume of data.
This is too large to transfer to the Audit workstation so a process running on the Audit server will filter this
data for records containing the required PAN.

The workstation application will accept a single PAN number (either in clear or hashed form) as input,
calculate the NBX journal hash values and retrieve the relevant NBX journal streams. Once the NBX
Audit Tracks have been retrieved and seal checked, the application will initiate an Audit Server process to
filter the retrieved files for entries containing the required PAN. The workstation application will monitor
the progress of the server process and report its success or failure back to the end user.

The server based application will perform a simple line based search for occurrences of the PAN string in
the retrieved Audit Tracks. It will produce an output file on the Audit server, containing records that
contain the PAN string. The auditor may then copy the output file back to the Audit workstation, reformat
it as appropriate, and submit it to the requestor.

If a hashed PAN was originally entered as input, then only records that contain the hashed version of the
PAN can be located. If a clear text PAN is entered, which is expected to be the norm, the search will
locate records which contain either the clear text form (originating from pre CP4305 audit tracks) or the
hashed form.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 33 of 42
FUJ00123758
FUJ00123758

Audit Data Retrieval High Level Design

oO
FUJITSU COMMERCIAL IN CONFIDENCE

6.1.7. PAN Hash Calculation

The Audit workstation contains a function than enables the user to calculate the CP4305 PAN Hash value
for a given clear text PAN. This is utilized by the Audit workstation when searching for transactions which
relate to a specific PAN.

The user will be able to invoke this function as a free standing feature within the workstation application
or implicitly when querying NBX data using a full clear text PAN (see section 6.1.6.3.)

The PAN hash function is specified in HNG-X Crypto Services High Level Design (DES/SEC/HLD/0002)
§3.2.1.

The PAN hash calculation is performed using the HNG-X Crypto API (See DES/SEC/IFS/0001). This API
will be utilized directly from the Audit Workstation Visual Basic applications.

6.1.8 PAN Decryption

The Audit workstation contains a freestanding feature than enables the user to decrypt a single encrypted
PAN. This is utilized when query has returned Hashed PAN & encrypted PAN data, and it is necessary
to positively identify the PAN in question.

The PAN decryption function is performed using the HNG-X Crypto API (See DES/SEC/IFS/0001). This
API will be utilized directly from the Audit Workstation Visual Basic applications. The Crypto API requires
access to an Attalla NSP, which is locally connected to the Audit workstation subnetwork.

The Audit user may need to decrypt PANs which go back up to seven years, thus the key server must
contain up to seven years of historic PAN decryption keys.

Usage of the PAN Decryption feature will be included in the Audit server audit trail.
Note that it is not possible to turn a Hashed PAN back into clear.

6.1.9 Data Delivery

Data will be delivered to the requester by CD-R media. CDs will be but using the standard Windows XP
CD writing feature.

The user must ensure that the Seal Checking has completed successfully before extracted data is further
processed and/or passed to the requestor.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 34 of 42
FUJ00123758
FUJ00123758

Audit Data Retrieval High Level Design

oO
FUJITSU COMMERCIAL IN CONFIDENCE

6.2 Distributed Application Services

Figure 6 show the distributed application interfaces between the components of the Audit system

Centera

EMC Centera API EMC Centera API

‘Tivoli Cluster Lookup),

HTTP
f cay
Windows File Windows File Windows fief
Sharin Sharing I =
Sharing 9 :

KING Server Audit Server Audit Server

SQL Server Database Connection

Key Server Windows File Sharing

Client API

Atalla API Atalla NSP

Audit Workstations

Figure 6 — Audit Distributed Application Interfaces

The ARQ SQL Server 2000 database is used to support client-Server communication between the Audit
workstations & servers. Service requests are initiated by writing to interface tables in the database.
The client monitors the progress of server actions via status tables in the database.

The Audit servers access the EMC Centera units using the EMC Centera API.

The Audit workstations access the Atalla NSP and the KMNG Server using the HNG-X Crypto API (See
DES/SEC/IFS/0001). .

The Tivoli cluster lookup service is used to determine which Correspondence Server cluster a particular
branch belonged to at a particular point in time. This is only relevant to Horizon Riposte transactions, but
access to this data must be maintained for the 7 years that Riposte data exists in the Audit archive. The
cluster lookup service will be retired once Sysman2 components are removed from the solution. At this
point the cluster look up data will be imported into the Audit system and accessed directly.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 3 of 42
Audit Data Retrieval High Level Design
COMMERCIAL IN CONFIDENCE

FUJITSU

FUJ00123758
FUJ00123758

6.3 Networking Services

The Audit Server will use the standard HNG-X network services.

a predictable set of ports.

Branch database

Local data centre network

Centera

Local data centre

network Tivoli Cluster Lookup

Local data \
centre network

Figure 7 shows the networked
interfaces between components of the Audit system. All application services utilise TCP and/or UDP on

Local data centre
network

Inter data
centre link

Local data
centre network

Local data
centre network’

KMNG Server Audit Server Audit Server

BRAO1/LEWO2 link to

BRAOI/ \ data centres

LEWO02 link to
data centres ~

‘Audit Workstations

Figure 7 — Network Services

BRA01/LEWO2 local
network

Atalla NSP

The Audit system can move large volumes of data around the network. On a peak day, it is estimated
that the audit system will gather 10 GB of data at the active data centre. This data will flow as follows

1. Local data centre network - HNG-X systems ®» Audit server 1
2. Local data centre network — Audit server 1 > Centera 1
3. Inter data centre link — Audit server 1 > Audit server 2
4. Local data centre link — Audit server 2 > Centera 2
In addition, during Hydra, the following additional flows will exist
1. Bootle/IRE11 link - Horizon systems > Audit server 1
2. Bootle/IRE19 link - Horizon systems » Audit server 2
3. Inter data centre link — Audit server 2 > Audit server 1

©Copyright Fujitsu Services Ltd 2070 COMMERCIAL IN CONFIDENCE

Ref:
Version:
Date:

Page No:

DES/APP/HLD/0029
V2.0

11/06/2010

36 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

During Hydra, a peak load of 35GB per day will flow from Bootle > IRE11 and Wigan > IRE19. The
volume of this traffic will decrease as the HNG-X counter rollout progresses.

The bulk of Audit data (by volume) is generated by the Branch database. This data will be made
available to the audit system by an overnight extract taken shortly after 01:00. Thus the audit system will
utilise the local network quite heavily during this period.

Replication of audit data must take place when the audit system is quiescent. This is scheduled to take
place after the Audit backups have completed. During this period the audit system will be heavily utilising
the inter data centre link. This may require some traffic management on the network to avoid the Audit
replication traffic flooding the link.

If the Audit server is in recovery mode (i.e. it has been out of action for a period of time) it will put a
considerably higher load on the network than stated above.

6.4 Platforms

The Audit servers run in an Active/Active configuration. They are housed in the BladeFrame.
The Audit servers will run Windows Server 2003 Standard Edition.

The Physical Platform Design for the Audit server is HNG-X Audit Server (ARC) - Physical Platform Design
(DES/PPS/PPD/0035)

New Audit workstations will be provisioned running Windows XP.

The Physical Platform Design for the Audit Workstations is HNG-X Audit Workstation (AUW) - Physical
Platform Design (DES/PPS/PPD/0037)

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 37 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

7 Systems Management

7.1. Monitoring

The Audit applications use standard Windows event logging to record application events. Events are
classified using the standard mechanism and are marked as Information, Warning or Error as
appropriate.

The SYSMAN3 Event Management system is responsible for analysing this event stream and taking
suitable action.

The Audit gathering system automatically manages the availability of free local disk space to hold
temporary copies of Audit Tracks. If the amount of free disk space falls below a given threshold, the
gathering system will suspend until sufficient free space becomes available. This is normally an
automatic process as other audit processes secure Audit Tracks and delete the local copy.

7.2 Schedule Requirements

The audit system is dependent on the HNG-X system scheduler to manage its operation. The scheduler
is used to:

e Detect fatal application failures & raise alerts

e Schedule (run and shutdown) individual Audit application processes
« Manage the cross campus replication of Audit Tracks

e Manage the Audit server backups

e During Hydra, manage the failover state of audit gathering processes

The Audit system schedule requirements are fully documented in Audit Server Schedule Design
(DEVANF/AON/0009)

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

Page No: 38 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

8 Application Development
The Audit retrieval system is developed using the following development tools and technologies
¢ Microsoft Visual Studio 6.0
o Visual Basic 6.0
o Visual C++ 6.0
e Microsoft SQL Server 2000
e Windows Scripting Host
e Microsoft Internet Explorer
« Microsoft Office 2003
e ActivePerl V5.10

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 39 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

9 System Qualities

9.1 Availability

The gathered Audit data held on an Audit Server is held on a RAID disk configuration so that a single disk
failure will not result in loss of the data.

A given ARQ may be processed on either Audit server. Thus temporary failure of a single Audit server
will not significantly impact the processing of ARQs.

The Audit system is unavailable for retrieval purposes while the audit system backups are taking place.
These backups are scheduled into an overnight slot.

Multiple Audit Workstations are provided to mitigate against the loss of a single workstation. The primary
Audit workstations are sited at BRAO1; alternative workstations are available at LEW02.

9.2 Usability

Where users are required to interact with the Audit Server such interactions are carried out, from the
Audit workstation. The Audit Extractor Client is used to initiate retrieval of Audit tracks and to submit
queries for execution. Standard facilities provided by COTS software including Windows XP standard
tools are used to access the Audit server Userarea share.

Some of the activities require relatively long elapsed times, e.g. recovery of Audit Tracks from secure
storage, to complete. The Windows based facilities provided on the workstation allow other activities to
be progressed while waiting for the longer term ones to complete.

The HNG-X Audit Extractor Client will allow multiple queries, each associated with a different ARQ, to be
submitted from a single workstation.

9.3. Performance

The implementation must enable the ARQ turnaround times defined in Security Management Service -
Service Description (SVM/SDM/SD/0017) to be met.

Clearly the broader the date range the larger the extraction is likely to be. Additionally, more complex
extraction criteria will have a greater impact on performance.

The file identification process is largely a manual process and will rely on the translation of the end-user
view of the data to an Audit View of the data, which is based upon a logical grouping of data using Audit
Points. If a large number of files have to be retrieved, which potentially contain the information required,
it may take a long time to locate the exact file(s) wanted.

When retrieving and querying branch transaction data, the time required to process a given ARQ should
not exceed that required by the equivalent Horizon feature. This is necessary to achieve ARQ
turnaround targets.

The interface to the HNG-X branch transaction data Query Processor will be integrated into the Audit
extractor client, this will minimise the number of user interactions required to fully process an ARQ which
relates to PO branch transaction data.

Staff availability may impact on the achievement of ARQ turnaround times.

Access to the Audit Atalla NSP will be manually initiated and infrequent. There are no performance
constraints.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 40 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

9.4 Security

Data retrieved using the Audit system is often used for litigation support purposes and is presented as
evidence in court cases. A number of measures are necessary to ensure the integrity of the data
generated by the retrieval process.

As Audit tracks are retrieved from the archive, they are seal checked (by re-application of the MD5
message digest function) to ensure that the source data has not been tampered with while it was stored
in the archive.

Only authorised users may access the Audit workstation applications. Authorised users are required to
log on to the workstation using two factor authentication & the HNG-X Identity Management system. An
Active Directory group named AUDIT_USER will be created with the rights required to utilise the
workstation applications. Authorised users will be added to this group.

User Log/On events are included in the Windows event log. Users are allocated to a specific role which
enables them to access the Audit databases.

All retrievals of audit data are performed using the Audit Extractor Client, and all such user actions are
themselves audited. It is not possible for users to access the archive by any other means.

Audit workstations and Atalla NSPs are located in secure areas. Only authorised users are given
physical access to these areas.

The standard HNG-X Anti-Virus solution (Sophos) will be installed on the Audit workstation. See HNG-X
Anti Virus High Level Design (DES/SEC/HLD/0011).

9.5 Potential for Change

When detailed Audit Information Requests and Audit Reports are specified by, POL and POL Clients,
detailed analysis can be carried out to determine if specific additional extraction and filtering facilities
need to be developed for other (non branch transaction) audit data

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 41 of 42
FUJ00123758
FUJ00123758

oO Audit Data Retrieval High Level Design
FUJITSU COMMERCIAL IN CONFIDENCE

10 Migration

Further details of Audit system migration, including details of the changes required at each migration
phase, is given in Audit Data Collection & Storage High Level Design (DES/APP/HLD/0030)

10.1 Audit Workstations

New Audit workstations will be provisioned using Windows XP in line with the HNG-X platforms
architecture requirements.

The existing Audit client applications must be ported to run under the Windows XP environment. It is
expected that the majority of client applications will run unchanged on Windows XP. This will entail an
initial development team exercise to check the compatibility of the existing client applications with
Windows XP & to identify & resolve any issues encountered.

A new workstation build must be developed based on Windows XP, Microsoft Office 2003 and the
updated applications.

Two of the current Horizon workstations will be retained at least until all Horizon counters have been
migrated to HNG-X. This is in case any existing ARQs require further work (say for a court case) and will
enable the original tool set to be used. This introduced a requirement that Riposte services continue to
be available on the HNG-X Audit server until all Horizon Audit workstations are decommissioned.

It will be the responsibility of the users of Audit workstations to backup any data that they may have
stored on the workstations prior to migration of the platforms.

10.2 Audit Servers

The Audit servers will be upgraded to Windows Server 2003 in line with the HNG-X platforms architecture
requirements.

The existing Audit server retrieval application components must be ported to run under the Windows
Server 2003 environment. It is expected that the majority of such applications will run unchanged on
Windows Server 2003. This will entail an initial development team exercise to check the compatibility of
the existing client applications with Windows Server 2003 & to identify & resolve any issues encountered.

At the end of the Hydra phase of the migration process the data held within the Tivoli cluster information
service will be imported into the Audit system ARQ database. See section 6.1.4.6.

‘©Copyright Fujitsu Services Ltd 2010 COMMERCIAL IN CONFIDENCE Ref: DES/APP/HLD/0029
Version: V2.0
Date: 11/06/2010

PageNo: 42 of 42