Unit tests can only be done in sequence.

Each unit test currently uses the same test db and port. So the tests have to be run in sequence with the -i flag.
This also introduces the problem of leftover data in the test db that could possibly make other tests fail down the line.