Return Home

Course Details

Course Description

Can you afford to have your core enterprise systems meltdown under load? SOA technology is widely promoted as a fundamental software infrastructure used in the creation of agile, scalable, robust, flexible, and adaptable enterprise systems. SOA is supposed to leverage business value from legacy systems, facilitate re-use, and reduce costs of total lifetime ownership. But… how do know if your new SOA implementation will perform adequately with your current business loads, and will it scale-up and cope with growing future business loads? Established technologies for performance analysis and system load testing do not work well when faced with complex, distributed, heterogeneous compositions of software services sourced from many different vendors and geographical locations. This NICTA course presents a novel approach, based on Service Oriented Performance Modelling, to understanding and managing performance-related risks in an SOA environment.

Target Audience: Program Managers, Enterprise Architects, Developers, Testers.

Brief Course Outline

Day 1 is intended for program managers and enterprise architects who are responsible for the development and deployment of SOA initiatives in their organisation. This course component provides a comprehensive overview of the business risks associated with major SOA projects. It emphasises the importance of early management and thorough analysis of critical performance attributes such as scalability, reliability, availability, and maintenance of SOA implementations. Day 1 provides an introduction to a performance modelling and simulation technology used to predict, analyse, and configure SOA implementations. Example performance models of enterprise SOA's will be presented, and simulations of system behaviour will be demonstrated. Using the predicted performance results from these simulations, the application and benefits derived from this technology at both project and enterprise levels of an organisation, are explained. Course participants have an opportunity to gain hands-on experience with the example performance models and run demonstrations on their laptops.

Day 2 provides further practical experience with the performance modelling and simulation technology introduced in Day 1. Course participants will be instructed in the techniques of constructing and calibrating performance models, and learn how to build predictive models of realistic SOA implementations. The simulation tools allow performance issues to be identified and analysed in the SOA implementation under varying business loads. Response times and throughputs can be explored under extreme loads and system "meltdown" scenarios can be identified. Using this information, the implementations can be re-architected to optimise performance, and system capacity can be determined. Participants will also learn how the SOA modelling technology can be applied to non-SOA implementations.

Pre-requisite: For Day 1 attendance, equipment is not essential, however, it is required by those participants who want to run demonstrations on their own laptop. Course participants who continue with the practical components of Day 2 will require a laptop running MS-Windows (XP or Vista). The instructors will install a demonstration copy of the modelling and simulation tools onto this laptop for completion of the practical exercises.

Assumed Knowledge: For Day 1, course participants are likely to have some years of experience in IT development and IT project management, but the background knowledge is not specific to any particular vendor technology. Some familiarity with basic SOA concepts is an advantage. The practical activities in Day 2 assume that participants have significant career experience in software development, software architectures, and/or system testing.

Course Outcomes

At the conclusion of Day 1, course participants will have gained an understanding of the kinds of business risks, and in particular performance risks, associated with SOA initiatives; and participants will know how to develop a strategy to address these performance risks in their organisation. For those participants who continue the course and complete Day 2, they will have gained practical experience with a performance analysis and prediction technology. A demonstration copy of the modelling and simulation tool will be provided so that further experimentation on the example SOA implementations can be done after course completion.

Topics Covered:

Day 1: Introduction and Overview of Performance Modelling

Performance Modelling

• SOA concepts and technologies
• Business risks associated with SOA initiatives
• The business case for performance modelling
• Use of performance modelling for risk mitigation
• Performance modelling in the software development lifecycle
• Performance drivers and requirements
• SOA performance introduction
• Performance metrics
• Performance testing
• Performance data for model parameterisation and validation
• Service Oriented Performance Modelling: overview of the method, steps, and tools
• Example SOA performance modelling problem and solution demonstrated

Day 2: Techniques and Tools for Service Oriented Performance Modelling

Using two detailed case studies, we cover the following topics:

• Using architectural artefacts to discover and model the SOA architectural features (services - simple and composite, workflows, deployment, server information, performance data)
• Modelling service compositions, and performance model composition (interchanging and reuse of model components)
• Modelling service orchestration mechanisms (e.g. workflow engines and Enterprise Service Buses)
• Virtualisation and its effects on performance
• Using performance modelling in establishing SLAs
• Exploring workload, architectural, deployment and resourcing alternatives
• Applying SOPM tools and techniques to non-SOA implementations

How to Register

To register for this course fill out this registration form and scan and email it to industryeducation@nicta.com.au or fax it to +61-8-8343-8711 or click on the "Register Online" button next to the course name on the Upcoming courses page.

Cancellation policy

At least four weeks notice is required for cancellation of a place in a short course for full reimbursement. If cancellation is later than 4 weeks then the place can either be given to another person or the registrant can be provided with a credit towards other NICTA training.

For further information please contact:
Anne-Marie Eliseo
Industry Education Manager
phone: +61-8-8343-8710
email: industryeducation@nicta.com.au