Managing Test Data in Enterprise Environments
Managing test data is hard. Managing test data in complex, distributed, poorly documented, and organization-spanning enterprise software is substantially harder. In enterprise systems test data is convoluted, hidden, and hard to synthesize, the test environments it flows through are shared and unstable, and the test scenarios dependent on it are long and complex. These challenges combine to cripple many test automation efforts in enterprise software. In this talk Blake Norrish will describe what an optimal test data management strategy looks like, how the nature of enterprise software prevents this, then show how the concepts of data sourcing, data lifecycle management, data location, architectural decomposition, and hermetic sealing can be used to help mitigate these challenges. With the necessary understanding of and investment in test data management, deterministic test automation should be possible even in the most complex enterprise software.