A test manager executed by the server computer analyzes the results of the tests performed by the client computers, and determines which tests have passed or failed during execution, whether the tests failed because of an application or other error, and which tests should be re-executed by the same or different client computer.