Synthetic Test Data Generation from Test Data Automation
High-speed, fully customizable test data generation. Test Data Automation builds realistic test data using accurate models of production data or files, creating complete and compliant data for test environments.
Rigorous testing requires complete data
Rigorous testing depends on realistic test data with which to execute every test, including negative scenarios and outliers. However, test data for many organisations still means masked copies of production data: large data sets that are typically low variety. The production data is drawn from the everyday and “happy path” scenarios exercised by typical users, and so contains only a fraction of the combinations needed for rigorous testing. This leaves systems exposed to costly and damaging defects.
QA or Ops teams might manually create test data, having searched through large data sets for the combinations they need. However, the data must be consistent across the numerous databases and files involved in the system under test. Manually generating data is therefore slow and laborious, while numerous automated test failures arise from incorrect test data.
Automatically generate test data for every possible test
Using Test Data Automation, QA or Ops teams can build data models automatically from a range of sources, retaining the relationships that exist across tables and columns for realistic test data generation. Easy-to-use parameters makes data definition quick and simple, while business rules can also be specified quickly. Soft relationships can additionally be added to the synthetic data, generating realistic, production-like data for rigorous testing.
This three part video series demonstrates how to generate high quality synthetic test data using Test Data Automation. Watch the example of an SQL eCommerce database to see how:
-
Accurate data models of existing data can be built automatically, pointing VIP to a range of database and file types. This includes SQL Server, Oracle, My SQL, Postgres, and MS Access.
-
The automated data modelling produces an easy-to-use spreadsheet for test data generation, retaining the existing and custom defined relationships.
-
A “fill-in-the-blanks” approach then allows you to specify dynamic or static test data for each table, using over 500 custom data generation functions to create complete test data with which to execute every test.
-
Existing data can be incorporated into the automated test generation, while event hooks reflect complex business logic accurately in the synthetic data, for realistic testing that matches your system under test exactly