Also detect errors in other parts of the tests

[marceloleitner]

We are only checking for error when the test itself executes, but we should also catch errors while preparing the environment for the test.