Business Technology Optimization in Practice: System Validation
Fabless Semiconductor Company
Limited code coverage and implementation preparation led to unplanned costs and implementation risks
APT Quality, APT Performance
- Performed Tests in 49% less time
- Saved Client nearly $1,000,000 on implementation
- Application was much more reliable upon implementation
- Set client up for future success with best practices of APT Methodology
A top tier Fabless Semiconductor company called on their long standing relationship with the Hexaware leadership team to assist with their Oracle 10.7 to 11i upgrade. From prior experience, the client knew they could test only about 70% of the core functionality of the system under standard manual methods, but they needed the new system to be up and running in a short period of time. The client felt that a lack of code coverage during the testing of the implementation was the cause of many of the problems they had faced in the past. They looked to Hexaware to help provide a solution that would improve their testing code coverage while still allowing them to meet an aggressive, bimonthly timetable for new module deployment.
Instead of proposing a simple test plan, the Hexaware leadership team shared with the client the vision of a new strategic initiative called Business Technology Optimization. They demonstrated how putting the focus on quality at the beginning of an application’s lifecycle could dramatically improve what was known to be a painful and costly process. Simultaneously, they presented a methodology for maximizing the value of the client’s entire IT ecosystem. Impressed with the plan and the team’s expertise in understanding business requirements as well as the technical issues, the client engaged Hexaware to embark on the first step of delivering on the promise of Business Technology Optimization.
The first phase of ensuring that the system achieved its optimal level of performance was to incorporate automated validation into the process. The Hexaware team used the industry’s most advanced tools to thoroughly examine all of the code of the core Oracle 11i functionality. The firm’s proprietary APTT Methodology added a discipline to the process by ensuring that all validation activities were based on real world business processes.
Through the course of the engagement, the Hexaware team created 325 scripts and ran on average twelve positive and negative tests per script to fully exercise the system. This equated to nearly 4000 individual tests that covered 95% of the core Oracle 11i functionality. In addition to putting the Oracle application through its paces, they also set the client up for success in future optimization efforts. The team created a re-usable, reliable automated regression test library to guarantee that all previous functions would work in conjunction with the new system. In fact, the client was able to make use of this regression test library to completely test the Oracle 11i implementation in only 4 days.
Once the upgrade was completed, Hexaware extended the APT Methodology through the release of the three modules that were to be added on: HR, Financials, and Order Management. In each case, the team was able to gain efficiencies through the use of a component based approach. This generated economies of scale and made it possible to create performance tests that were large enough in scope to eliminate the risk that the system couldn’t handle the estimated load for each phased release.
APT in Action
- Design Validation Strategy
- Implement Test Plan
- Identify and Eradicate Failures Early
- Experts in Application
- Experts in Integration
HP’s TestDirector and WinRunner
The effort from Hexaware’s delivery of the APT Methodology yielded benefits to the client almost immediately. The team was able to increase the amount of code tested by nearly 50% in 15% fewer days, just during the implementation phase. As the new components rolled out, the test coverage remained high while costs continued to drop. Furthermore, the team was able to identify and eradicate significant bottlenecks in the components before they were released into the production environment. Hexaware was even able to test their system 25% beyond their highest anticipated capacity so that unforeseen spikes in system usage would not cause problems.
Ultimately, the client saved nearly $1,000,000 on work that was accomplished in 49% of the time they would normally expect. But equally important was the improved reliability of the system that was the direct benefit of greater code test coverage. Too often, new technology offers a great deal of promise, but doesn’t deliver because of poor preparation. Thanks to Hexaware’s work this client got an impressive return on their IT investment more quickly and easily than they ever imagined.