Configure cluster to receive notifications via email

Setup cluster notifications for your GKE cluster

To use Pub/Sub to receive notifications about your GKE cluster, follow Receive cluster notifications.

Create new integration

To create a new integration, perform the following steps:

1. Go to the Application Integration page in the Google Cloud console.

   Go to Application Integration

2. In the navigation menu, click Integrations. The Integrations page appears.

3. Click Create integration.

4. In Integration name, enter a name for the integration you are creating.

5. If you have multiple regions enabled, a drop-down option will show for you to select a region for the integration. When only one region has been enabled, it will default to that region and the drop-down won't be displayed.

   Note: The Regions drop-down only lists the regions provisioned in your Google Cloud project. To provision a new region, Click Enable Region that is displayed at the bottom of the box.

6. Click Create to open the integration editor.

Add and configure Pub/Sub trigger

To add a Pub/Sub trigger to the integration, follow these steps:

1. In the integration editor, select Triggers to display a list of available triggers.

2. Click and place the Pub/Sub trigger element to the designer.

3. Click the Pub/Sub trigger element in the integration editor to view the trigger configuration pane.

4. Specify the topic the trigger should listen to in the Pub/Sub topic field in the following format:

   projects/PROJECT_ID/topics/TOPIC_ID
   

   Replace the following:

• PROJECT_ID: the Google Cloud project where your topic was created.

• TOPIC_ID: the ID of the Pub/Sub topic you created in Setup your environment.

  For example, if your project is named my-project and you named your topic gke-notifications, enter:

  projects/my-project/topics/gke-notifications
  

  To view all the available topics in your Google Cloud project, see the Topic name column in Topics. The Pub/Sub trigger configuration pane will automatically populate the Trigger Output field with a variable to hold the Pub/Sub message.

Add and configure the Data Mapping task

The Data Mapping task helps to extract relevant variables from the Pub/Sub notification. It takes the Pub/Sub message JSON file as input and extracts variables as output.

To configure the Data Mapping task:

1. Select Tasks to display a list of available tasks.

2. Click and place the Data Mapping element in the integration editor.

3. Click the Data Mapping task on the designer to open the configuration pane for the task.

4. To add an edge connection from Pub/Sub trigger to Data Mapping, hold the pointer over a control point on the Pub/Sub trigger, then drag a line to a control point on the Data Mapping task. The edge denotes the flow of control from the Pub/Sub trigger to the Data Mapping task.

5. Click Open Data Mapping Editor. The Data Mapping Editor lets you map input variables to the selected output variables, using the available transform functions. The output is then available as a variable to any other integration task or trigger. For more information about variables in Application Integration, see Variables.

   For this tutorial, follow these steps to create a mapping using the CloudPubSubMessage as input:

   1. Expand the CloudPubSubMessage JSON variable in the Variables list and drag the CloudPubSubMessage.data variable to the first Input row.
   2. To create a new output variable to hold the value of the transformed data, click the first row in the Output column. Then, do the following:
      1. In the Name field, enter message_data.
      2. From the Variable Type list, select Output from integration.
      3. From the Data Type list, select String.
      4. Click Create. The message_data variable appears in Output.
   3. Drag the CloudPubSubMessage.attributes variable from the Variables list to the second row in the Input column.

   4. Click the second row in the Output column to create a new output variable to hold the value of the JSON payload with the cluster attributes. Complete the following fields:

      1. In the Name field, enter attributes.
      2. From the Variable Type list, select None.
      3. From the Data Type list, select JSON.

      4. Select Enter a JSON schema from the JSON schema options, and paste the following payload:

         {
           "$schema": "http://json-schema.org/draft-07/schema#",
           "type": "object",
           "properties": {
             "cluster_name": {
               "type": "string"
             },
             "payload": {
               "type": "string"
             },
             "project_id": {
               "type": "string"
             },
             "cluster_location": {
               "type": "string"
             },
             "type_url": {
               "type": "string"
             }
           }
         }
         

      5. Click Create. The attributes variable appears in Output. Make sure the attributes variable looks similar to the following:

         

   5. Click the third row in the Input column and select attributes.cluster_name from the Variables pane. You might need to start typing attributes to see all 5 attributes defined from the schema entered in previous step.

   6. Click the corresponding row from the Output column to create a new output variable that'd hold the value of the cluster name attribute. Complete the following fields:

      1. In the Name field, enter cluster_name.
      2. From the Variable Type list, select Output from integration.
      3. From the Data Type list, select String.
      4. Click Create. The cluster_name variable appears in Output.

   7. Repeat steps e and f for project_id, cluster_location, type_url and payload.

   8. Close the Data Mapping Editor once your mapping is complete. Any changes will be autosaved.

Add and configure the Data Transformer (Script) task

The Data Transformer task takes variables extracted from the Data Mapping task as input and transforms it into a message body for end user consumption.

To configure the Data Transformer task:

1. Select Tasks to display a list of available tasks.
2. Click and place the Data Transformer (Script) element in the integration editor.
3. Click the Data Transformer task on the designer to open the configuration pane for the task.
4. To add an edge connection from Data Mapping task to Data Transformer task, hold the pointer over a control point on the Data Mapping, then drag a line to a control point on the Data Transformer task. The edge denotes the flow of control from the Data Mapping task to the Data Transformer task.

5. Click Open Data Transformer Editor. The Data Transformer Editor lets you write, edit, and evaluate custom Jsonnet templates to perform data mapping in your integration. The output is then available as a variable to any other integration task or trigger. For more details on this transformer task, see Data Transformer Task. For this quickstart, follow these steps to create a script:

   1. Click add_box Create in the Variables left panel. This will open a Create variable panel on the right side. Complete the following fields:

      • Name: Enter full_message.
      • Select Output from integration in the Variable Type drop-down box.
      • Select String in the Data Type drop-down box.
      • Click Create to create the variable and close the pane. The new full_message variable will appear in the Variables list.

   2. Select Script.

   3. Use the following example script as a template for the message body when notifications are sent from your cluster:

      local message_data = std.extVar("message_data");
      local cluster_name = std.extVar("cluster_name");
      local project_id = std.extVar("project_id");
      local cluster_location = std.extVar("cluster_location");
      local type_url = std.extVar("type_url");
      local payload = std.extVar("payload");
      
      {full_message: "Project ID: " + project_id + "\n" +
                    "Cluster Name: " + cluster_name + "\n" +
                    "Location: " + cluster_location + "\n" +
                    "Event Type: " + type_url + "\n" +
                    "Message: " + message_data + "\n" +
                    "Payload: " + std.manifestJson(std.parseJson(payload))}
      

   4. Close the Data Mapping Editor once your mapping is complete. Any changes will be autosaved.

Add and configure the Send Email task

To configure the Send Email task:

1. Select Tasks in the integration editor to display the list of available tasks.
2. Click and place the Send Email element in the integration editor.
3. To add an edge connection from Data Transformer task to Send email task, hold the pointer over a control point on the Data Transformer task, then drag a line to a control point on the Send email task. The edge denotes the flow of control from the Data Transformer task to the Send email task.
4. Click the Send Email element on the designer to open the configuration pane of the task. Set the following properties:
   1. To Recipient(s): Enter an email address. You will use this email to confirm the successful completion of the integration.
   2. Subject: Enter GKE Notifications.
   3. Body in Plain Text: Select the full_message variable created earlier in the Data Transformer task.
   4. The remaining options can be left in the default configuration.

Your changes to the properties are saved automatically and your canvas should look similar to the following:

Test integration

To test the new integration, go to the navigation bar at the top of your Integration diagram, and click Test. It will open a panel on the right side. You can use the following sample Pub/Sub message in the Add a json value section:

{
  "data": "Node pool projects/my-project/locations/us-central1-c/clusters/cluster-1/nodePools/default-pool is upgrading to version 1.29.1-gke.1589017.",
  "attributes": {
    "cluster_location": "us-central1-c",
    "cluster_name": "cluster-1",
    "payload": "{\"resourceType\":\"NODE_POOL\",\"operation\":\"operation-upgrade\",\"operationStartTime\":\"2024-05-07T20:09:35.528855756Z\",\"currentVersion\":\"1.28.7-gke.1026000\",\"targetVersion\":\"1.29.1-gke.1589017\",\"resource\":\"projects/my-project/locations/us-central1-c/clusters/cluster-1/nodePools/default-pool\"}",
    "project_id": "my-project",
    "type_url": "type.googleapis.com/google.container.v1beta1.UpgradeEvent"
  }
}

An email is sent to the email address that you have configured in the Send Email task after clicking the Test Integration button.

Publish your integration

After you test this integration successfully, click Publish for your integration to start processing the messages.

You can edit only a DRAFT version of the integration, and execute only the ACTIVE version of the integration. If you want to edit your integration, click Enable Editing on the integration designer page and then create a draft version. For more information, see Integration edit locks.