Test Data Management
In a financial business world dominated by DATA intensive applications there is constant need to collect, manipulate, store and retrieve of large amounts of data. In addition most of the applications being process oriented require analysis and implementation of complex business rules to the data.
Few examples are taxation, customs, excise application, mutual funds, Fund Management organizations, where the underlying data typically scales over time.
Given that the tolerance for faults and errors in such applications is very small, focus has to be bought on how to effectively use data for robust functional testing.
Mostly the data challenges are possible to manage from a manual testing perspective, but that becomes very cumbersome and inefficient overtime, especially for regression testing.
In order to automate the regression of such applications, usually there are 2 options:
- Automation tool and scripts build in the complex business rules and validation points in-line with the application. This implies that when tests are executed with multiple data sets, the values calculated by scripts should match the ones generated by the application.
- The other approach is to use Managed Data to test the functionality of the application. This approach proves more beneficial for data intensive applications where input data is changing frequently. It allows the functionality to be tested against a predictable outcome and insulates the application from external and internal dependencies.
- The focus of our discussion will be ‘Testing with Managed Data’ – framework details, prerequisites, pros and cons compared to other approaches.