monasca team mailing list archive

[Nicolas Seyvet <nseyvet@xxxxxxx>]

Hi,

Sorry if this is the wrong channel.

We are having a difficult time understanding the purpose of the following tests in the monasca-agent test:
-  TestKeystone.test_no_override
-  TestKeystone.test_override_values

First, the test suites fail when ran from the monasca-agent root directory:


github/monasca-agent$ nosetests -w tests
..................................................................................................................................................................................................................EE.
======================================================================
ERROR: Test setting values with no overrides works as expected
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/local/git/github/monasca-agent/tests/test_service_plugin.py", line 40, in test_no_override_
    processes = config['process']['instances']
KeyError: 'instances'
-------------------- >> begin captured logging << --------------------
monasca_setup.detection.service_plugin: INFO:  Configuring an http_check for the os-testr API.
--------------------- >> end captured logging << ---------------------

======================================================================
ERROR: Test overriding values using args works
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/local/git/github/monasca-agent/tests/test_service_plugin.py", line 59, in test_override_values
    processes = config['process']['instances']
KeyError: 'instances'
-------------------- >> begin captured logging << --------------------
monasca_setup.detection.service_plugin: INFO:  Configuring an http_check for the os-testr API.
--------------------- >> end captured logging << ---------------------

----------------------------------------------------------------------
Ran 213 tests in 12.726s

FAILED (errors=2)

When trying to fix that, the test code searches for a “testr” process, and some other processes that do not have a “cmdLine”, and we can not figure out what the code should do in either of those cases and/or how to “mock”/fake the testr process?

Any help is appreciated.


--

Med vänlig hälsning / Kind regards
___________________________________________________________
Nicolas Seyvet | Backend engineer | OP5
Mobile: +46 (0) 70 5640795
Email: nseyvet@xxxxxxx