Only use parameter_defaults for scaling down

We were using parameters for scaling down but parameter_defaults for
scaling up. The old parameters from scale-down still persisted in the
stack and took precedence over the scaled-up parameter_defaults, so the
scaled-up values didn't propagate.

We already update parameter_defaults in the deployment plan in that same
method, so all we need to do is remove the 'parameters' part.

Change-Id: I90a195a66b15c7c9057bfb415b6a730bbd9a939c
Closes-Bug: #1641142

tripleo_common/actions/scale.py

@@ -83,11 +83,6 @@ class ScaleDownAction(templates.ProcessTemplatesAction):
         fields = processed_data.copy()
         fields['timeout_mins'] = timeout_mins
         fields['existing'] = True
-        # TODO(rbrady): test against live overcloud to see if it is required
-        # to pass 'parameters' here to the stack update, or will storing the
-        # updated values in the plan and send them merged into the environment
-        # work as well
-        fields['parameters'] = parameters
 
         LOG.debug('stack update params: %s', fields)
         self._get_orchestration_client().stacks.update(self.container,

tripleo_common/tests/actions/test_scale.py

@@ -119,10 +119,4 @@ class ScaleDownActionTest(base.TestCase):
             environment=env,
             existing=True,
             files={},
-            timeout_mins=240,
-            parameters={
-                'ComputeCount': '0',
-                'ComputeRemovalPolicies': [
-                    {'resource_list': ['node0']}
-                ]
-            })
+            timeout_mins=240)