No need to merge heat templates with current one because we rebuild it on every deployment

Change-Id: I9a598aac6b6f3d53dae9d8b3d4870218b8babfe0

data/workflows/AspNetApps.xml

@@ -43,7 +43,7 @@
 		</update-cf-stack>
 	</rule>
 
-	<rule match="$.services[?(@.type in ('webServerFarm', 'aspNetAppFarm'))].units[?(@.state.hostname and not @.state.registeredWithLB)]"
+	<rule match="$.services[?(@.type in ('webServerFarm', 'aspNetAppFarm'))].units[?(@.state.hostname and not @.temp.registeredWithLB)]"
           desc="Units of web-farms services which have a hostname assigned but are not registered with LB">
 		<update-cf-stack template="LoadBalancer" result="outputs">
 			<parameter name="mappings">
@@ -54,7 +54,7 @@
 				</map>
 			</parameter>
 			<success>
-				<set path="state.registeredWithLB"><true/></set>
+				<set path="temp.registeredWithLB"><true/></set>
 				<set path="::uri">http://<select source="outputs" path="LoadBalancerIP"/>:<select path="::loadBalancerPort"/></set>
 			</success>
 		</update-cf-stack>

muranoconductor/commands/cloud_formation.py

@@ -110,6 +110,10 @@ class HeatExecutor(CommandBase):
         template, arguments = self._get_current_template()
         stack_exists = (template != {})
 
+        # do not need to merge with current stack cause we rebuilding it from
+        # scratch on every deployment
+        template, arguments = ({}, {})
+
         for t in self._update_pending_list:
             template = muranoconductor.helpers.merge_dicts(
                 template, t['template'])

tests/conductor/test_heat_commands.py

@@ -136,7 +136,6 @@ class TestHeatExecutor(unittest.TestCase):
                 'arg1': 'arg1Value',
                 'arg2': 'arg2Value'},
             template={
-                'instance': {},
                 "testName": {
                     "testKey": "testValue"
                 }