e5b8ebde85
There appears to be a race condition related to language parsing that is being uncovered when we mock open around the arg parse code. This highlights that we're relying to open mock instead of mocking a function that uses open. Rather than rewrite all the tests, let's reduce the scope of the open mock to the action execution instead. This only really applies to train and older because we rewrote many of these actions to ansible executions instead. Change-Id: I7f7b25aa76bef692991200ba621a536f8868200a Closes-Bug: #1909118 |
||
---|---|---|
.. | ||
config | ||
fixture_data | ||
v1 | ||
workflows | ||
__init__.py | ||
base.py | ||
fakes.py | ||
test_export.py | ||
test_overcloud_credentials.py | ||
test_plugin.py | ||
test_utils.py |