Brian Grindley
Capacity Planning Services Ltd.

This paper will demonstrate the techniques used by the author to size development environments and will discuss work to refine the algorithms used.

The cost of getting the sizing effort wrong is usually apparent in cases of overconfiguration, but where sites are underconfigured it is difficult to say what the cost is to the organisation. Brian will discuss how it may be possible to put an economic value on the cost of underconfiguration.

This case study examines the evaluation of UNIX (RS/6000, SUN, HP) infrastructure capacity alternatives within a large financial institution The focus of this paper is to examine what really takes place 'behind the scenes' during the assessment and evaluation of UNIX capacity alternatives. The case will review the technical objectives but will concentrate on understanding the socio-political dynamics which are present in all organisational situations. During the course of the case study it was discovered that there were two, distinguishable and separate evaluations taking place one formal and the other informal. By attending this session you will be better able to identify and understand the dynamics and implications of more than just the technical / capacity aspects of a typical IT evaluation.

Project Management in Practice
Kim Ko
IBM GSA

Often underrated and misunderstood, Project Management and Project Management Culture can contribute significantly to an IT organisation in financial and management terms. In this paper, the author shall discuss the concepts of project management, what it really means to have a project management culture, Consultancy Vs Project Management, Deliverable Vs Delivery, Pitfalls, Value Adds, etc., from a practical viewpoint within the infrastructure environment.

Performance measurement of commerical outsourcing arrangements has evolved from measuring primarily service delivey to measuring a number of key "partnering" drivers such as service quality, finance, relationship values and value add. The Balanced Scorecard provides the "dashboard" to setting the expectations of both parties and measuring the acheivement of those expectations. It transcends both the contract and service level agreement to combine the high subjective expectations with the monetary and service outcomes. This session will guide you through the process of developing an outsourcing Balance Scorecard in context of different forms of outsourcing and "partnering" arrangements. It is the new way to select the best service provider and to ensure long term success is achieved.

IT on the Move...
Pierre Louys
State Rail of NSW

State Rail of NSW is aiming at relocating a number of its business units from various locations in Sydney CBD and consolidating them into three (3) buildings (Central, Xerox house, and CityRail Centre). As part of the project, the IT division (IT Infrastructure) has been given the task to:

• decommission, relocate and commission Information Technology and Telecommunications (IT&T) infrastructure at CityRail Centre, Xerox, and Central buildings
• perform the migration of workforce computer equipment for an estimated 1,300 staff.

With the advent of e-systems, IT can no longer get away with the vertical silo views of Application Service. Business Units look at systems as seen by the end users. Systems Management must achieve the same view. This will allow IT to manage the e-systems from end-to-end, in a way that matchs the end user point of view.

Expanding Your IT Service Vision
Ian Jones
Australian Bureau of Statistics

In many organisations, SAS Institute's ITSV has been used to address the enterprise's needs for comprehensive warehousing and reporting on IT infrastructure performance. Following convention, the Australian Bureau of Statistics (ABS) purchased ITSV primarily to provide reporting on its complex IT infrastructure including data network and servers. Fundamentally, ITSV is a transactional-based data warehousing and reporting system, offering tailored solutions to capacity planners. Detailed analysis of ITSV architecture will show that administrators are armed with the ability to provide data warehousing services to non-conventional clientele including helpdesk management, cost recovery management and project managers. In this paper, the author reviews the fundamentals of ITSV, and how the system's flexibility allows administrators to expand their services. As a member of the ABS Network Management team, the author provides a case study of his experiences in expanding ITSV's portfolio to the benefit of the ABS.

Effective Gathering and Analysis of NT/2000 Performance Data
Craig Linn
University of Western Sydney, Nepean

NT 4.0 and its successor, Windows 2000, can provide a vast and rich array of system wide and application/service specific performance data. In the first half of this paper we examine: what performance data is available; where it is obtained from; and the variety of means available to access it, from low level API programming to very user friendly point and click GUI displays. In the second half we examine how you can analyse this data once you have it, and how best to go about integrating the data gathering and analysis phases. The above is couched within the context of commonly available tools, some provided with NT/2000 itself as standard, and others freely available on the Net.

This session covers security concepts for UNIX. In particular, it goes over some of the legal issues, covers the tools hackers use, the documents they use and how to secure your system against them. It also includes information on what to do when you are broken into. The intent of the session is not to teach attendees how to be hackers, but rather to give examples of how systems can be secured against them. The focus of the second half will be on UNIX. This material will be presented in two, back-to-back sessions.

SMP has a long and varied history. Only the developers could honestly describe it as “long and honourable”. There are some of us that remember the days of stage 1 & 2 sysgens; then came SMP4 then SMP/E. Various packaging solutions have also come, and in some cases gone - IPO, PDO, Custompac and the latest and greatest, ServerPac. Guaranteed to work - IBM set it up, IPL it, dump it to tape, and ship it to you with a handy little installer.

What could be better:

• any idiot can sell it
• any idiot can see the benefits and order it
• any idiot can install it
• any idiot can package thirdparty software to suit.

Monitoring HFS Performance with DFSMS 1.5
Clark L Kidd
Landmark Systems

Each new release of OS/390 seems to feature more components that rely on functions provided by UNIX System Services. Thus, the proper tuning of OS/390 UNIX is becoming a critical and non-trivial component of overall OS/390 performance. The heart of the OS/390 UNIX file system is the HFS (Hierarchical File System), and many OS/390 performance problems can be traced to poor decisions related to HFS data sets. The good news is that DFSMS 1.5 vastly improved HFS performance, and added some new controls for tuning these important data sets. The bad news is that most of these new controls have not been well documented and explained. This session will attempt to correct that oversight by explaining basic HFS concepts, exploring the new controls added by DFSMS 1.5, and providing some practical suggestions for improving HFS performance.

Workload Manager and CICS: New Options and New Insights
Steve Samson
Candle Corporation, USA

Since the announcement of Workload Manager (WLM) in 1994, the interaction between CICS and WLM goal mode has been a frequent matter for concern. Many who implement goal mode see the complexity of the CICS interface and the difficulty of exploiting response time goals; they give up and stay on velocity goals with no intent to change.

On May 16, 2000, IBM announced a series of new facilities to be available in the September 2000 time frame to overcome some objections to goal mode. Some optional changes to the way CICS operates in goal mode are among those new facilities.

This paper presents the new facilities as they apply to CICS as well as a new and less disruptive approach to goal mode migration and exploitation for CICS.

Some have said that batch tuning is a dying trade, there are other, new, exiting technologies that Capacity and Performance analysts should be dealing with. Why then is it still necessary?

Although batch tuning techniques are certainly 'mature' there are now many reasons why the batch window has to be squeezed. Businesses want their on-line systems available earlier in the morning and later at night, continuity of business / disaster recovery backups may be eating into the critical path, business growth may be excessively high or perhaps the batch processing was not well designed in the first place.

This paper documents a case study where a major international corporate bank had to reduce the end-of-year batch window '…by at least four hours' to allow Euro conversion processing to take place over 'le weekend'.

SANs are a hot topic this year. Everyone wants to build one but few people have achieved one to date. This paper focuses on one user's experiences on building a SAN over the last 18 months. Suncorp-Metway were a very early adopter of the technology. They saw the need to rationalise their storage environment and to achieve economies of scale. Their SAN has been achieved with a few teething troubles experienced along the way. Nevertheless, the results have been impressive. They have managed to integrate OS/390, HP-UX and Windows NT in a multi-vendor, multi-location environment, both for disk and tape. This paper discusses user experiences, how to avoid the pitfalls and what to expect when your "SAN Box" is up and running.

Is Your Organisation Ready for SAN?
Oscar Ernst
Qantas

The implementation of SANs (Storage Area Networks) in medium and large corporations that are running two or more different platforms presents enormous technical, strategic, organisational, political and cultural issues. It is up to vendors to provide solutions for the technical problems but it is up to the users to find answers for all of the other issues.

Considering that not all companies are the same, Storage Managers evaluating a SAN implementation should expect little or no help from vendors and/or external consultants when trying to evolve from the traditional "management by platform" model to the "management by function" model required by SAN.

This paper will try to identify most of these issues and suggest different approaches to tackle the problem.

Introduction to Storage Area Networks
Bruce Howarth
University of Technology, Sydney

Storage Area Networks (SANs) may be the Next Big Thing. This presentation will identify the business and technological drivers behind SANs and associated products such as Network Attached Storage (NAS), and will attempt to describe the current state of development in the field. Several products are available, and several groups are developing standards for aspects of SAN operation. I will try to show how these are related.

Practical SAN's
Andrew Tippett
Storage Technology
The acronym SAN has produced a frenzy of information by both vendors and analysts. This presentation cuts through this mass of information and uses real life Australian implementations of SANs in multi-vendor, heterogeneous environments to give you the facts on what has and can be achieved today, why you would do it, and the business reasons to move ahead with a SAN implementation.

Capacity Planning and Performance Management use a wide range of techniques. To the uninitiated some of these methods can appear to be a 'black art' employing 'smoke and mirrors'. This paper shows how a variety of techniques, from simple to complex, helped a site predict accurately the effect that employing a Remote Mirroring technique for Continuity of Business had on its batch window.