fix(uat): disable T17 on Windows and fix T18

aws-greengrass · Jul 27, 2023 · 95a875c · 95a875c
1 parent f3f2d38
commit 95a875c
Showing 1 changed file with 15 additions and 8 deletions.
diff --git a/uat/testing-features/src/main/resources/greengrass/features/ggmq-1.feature b/uat/testing-features/src/main/resources/greengrass/features/ggmq-1.feature
@@ -1066,8 +1066,8 @@ Feature: GGMQ-1
       | v5     | paho-python | aws.greengrass.client.Mqtt5PythonPahoClient | client_python_paho.yaml |
 
 
-  @GGMQ-1-T17 @OffTheNetwork
-  Scenario Outline: GGMQ-1-T17-<mqtt-v>-<name>: As a customer, I can configure IoT Core messages to be forwarded to local MQTT topic, with Greengrass in offline state
+  @GGMQ-1-T17 @OffTheNetwork @SkipOnWindows
+  Scenario Outline: GGMQ-1-T17-<mqtt-v>-<name>: As a customer, I can configure IoT Core messages to be forwarded to local MQTT topic, with Greengrass restarted in offline state
     When I create a Greengrass deployment with components
       | aws.greengrass.clientdevices.Auth        | LATEST                                  |
       | aws.greengrass.clientdevices.mqtt.EMQX   | LATEST                                  |
@@ -1140,8 +1140,9 @@ Feature: GGMQ-1
     And I verify greengrass-cli is available in greengrass root
 
     When I set device mqtt connectivity to offline
+    # Not works on Windows due to https://github.com/aws-greengrass/aws-greengrass-testing/issues/208
     And I restart Greengrass
-    And the greengrass log on the device contains the line "com.aws.greengrass.mqttclient.AwsIotMqtt5Client: Failed to connect to AWS IoT Core" at least 3 times within 2 minutes
+    And the greengrass log on the device contains the line "com.aws.greengrass.mqttclient.AwsIotMqtt5Client: Failed to connect to AWS IoT Core" within 2 minutes
 
     And I update my local deployment configuration, setting the component aws.greengrass.clientdevices.mqtt.Bridge configuration to:
     """
@@ -1161,8 +1162,11 @@ Feature: GGMQ-1
 
     And I set device mqtt connectivity to online
     And the greengrass log on the device contains the line "com.aws.greengrass.mqttclient.AwsIotMqtt5Client: Successfully connected to AWS IoT Core" at least 2 times within 1 minutes
-    And the greengrass log on the device contains the line "com.aws.greengrass.mqtt.bridge.clients.MQTTClient: Connected to broker" at least 2 times within 1 minutes
-    And I wait 30 seconds
+    # On Linux EQMX run in docker and use different IP network, when we block MQTT we also block MQTTClient of Nucleus.
+    # But on Windows EQMX run on the IP of local machine and connection of Nucleus to EQMX is not broken, so we miss secondary "Connected to broker"
+    #  Finally we replace with wait and in some cases scenario will failed
+    # And the greengrass log on the device contains the line "com.aws.greengrass.mqtt.bridge.clients.MQTTClient: Connected to broker" at least 2 times within 1 minutes
+    And I wait 65 seconds
 
     # Is critical to do discover after restart?
     When I discover core device broker as "localBroker" from "localMqttSubscriber" in OTF
@@ -1261,7 +1265,7 @@ Feature: GGMQ-1
     And I verify greengrass-cli is available in greengrass root
 
     When I set device mqtt connectivity to offline
-    And the greengrass log on the device contains the line "com.aws.greengrass.mqttclient.AwsIotMqtt5Client: Failed to connect to AWS IoT Core" at least 3 times within 2 minutes
+    And the greengrass log on the device contains the line "com.aws.greengrass.mqttclient.AwsIotMqtt5Client: Failed to connect to AWS IoT Core" within 2 minutes
 
     And I update my local deployment configuration, setting the component aws.greengrass.clientdevices.mqtt.Bridge configuration to:
     """
@@ -1281,8 +1285,11 @@ Feature: GGMQ-1
 
     And I set device mqtt connectivity to online
     And the greengrass log on the device contains the line "com.aws.greengrass.mqttclient.AwsIotMqtt5Client: Connection resumed" at least 2 times within 2 minutes
-    And the greengrass log on the device contains the line "com.aws.greengrass.mqtt.bridge.clients.MQTTClient: Connected to broker" at least 2 times within 1 minutes
-    And I wait 30 seconds
+    # On Linux EQMX run in docker and use different IP network, when we block MQTT we also block MQTTClient of Nucleus.
+    # But on Windows EQMX run on the IP of local machine and connection of Nucleus to EQMX is not broken, so we miss secondary "Connected to broker"
+    #  Finally we replace with wait and in some cases scenario will failed
+    # And the greengrass log on the device contains the line "com.aws.greengrass.mqtt.bridge.clients.MQTTClient: Connected to broker" at least 2 times within 1 minutes
+    And I wait 65 seconds
 
     When I discover core device broker as "localBroker" from "localMqttSubscriber" in OTF
     And I label IoT Core broker as "iotCoreBroker"