Instead of defining a Config class for functional tests, this patch integrates it with Marconi's base test suite and replaces the use of ConfigParser with olso.config. The patch doesn't change current behavior and gives the possibility to use external configuration files for both the functional test and Marconi. blueprint refactor-system-tests Change-Id: I60de6e199182881e945a24525c7c4f2d6937256f
Marconi Functional Tests ====================
Marconi's functional tests treat Marconi as a black box. In other words, the API calls attempt to simulate an actual user. Unlike unit tests, the functional tests do not use mockendpoints.
Running the Functional Tests ------------------------
Setup a Marconi server. Refer to the Marconi README on how to run Marconi locally, or simply use an existing server.
Install functional tests dependencies. :
pip install -r functional-test-requirements.txtcd to the marconi/tests/functional directory
Copy marconi/tests/etc/functional-tests.conf-sample to one of the following locations:
~/.marconi/functional-tests.conf /etc/marconi/functional-tests.confUpdate the config file to point to the Marconi server you want to run the tests against
If leaving keystone auth enabled, update system-tests.conf with a valid set of credentials.
Now, to run the sytem tests, simply use the nosetests commands, e.g.:
Run all test suites: :
nosetests --tests tests.functional -v
Adding New Tests
Add test case to an appropriate test case file: :
queue/test_queue.py messages/test_messages.py claim/test_claims.py