The bank's original product-testing team generally used offline testing. However, as the bank product line expanded the testers' workload increased, resulting in frequent overtime. Employees complained, and the quality of testing deteriorated.

The client used two performance-testing tools: LoadRunner and JMeter. However, the increasing performance testing demand meant there was inefficient script management, test execution, problem identification, resource monitoring and result analysis. Meanwhile, based on the current tools and methods, only a limited number of transactions could be selected for pressure testing. Consequently, the test results were biased. They could not reflect and identify the actual performance problems. To solve such issues, it was necessary to set up a performance-diagnosis platform.