A computer implemented method of automatic software testing. Status information of a software test running on a test system is stored to a common information point. The common information point is typically on a computer system independent from a computer system running the software test. An operating system on the test system is reinitialized. The operating system may be changed. The common information point is queried to determine the status information and the software test is resumed. The testing may be resumed at a point immediately after a last completed test operation.