In this hands-on lab, we will introduce ourselves to working with Google Cloud Deployment Manager. We will start with a simple single-VM deployment, move on to a more complex deployment with multiple dependencies, and finish by deploying a group of imported template configurations in a single deployment.
Learning Objectives
Successfully complete this lab by achieving the following learning objectives:
- Download configuration files used for this lab
We will perform the majority of our activities in Cloud Shell.
- From the web console, open Cloud Shell from the top-right
Download the configuration files we will use for this lab by copy/pasting the following command:
gsutil -m cp -r gs://acg-gcp-labs-resources/deployment-manager-basic/* .
Note: The period at the end of the command is required.
- Create your first single VM deployment
Before we create our first deployment, we must first edit our configuration file to reflect the current project.
From Cloud Shell, open the editor window by clicking the pencil icon in the top-right.
In the new window that appears, click on
vm-web.yaml.In the
vm-web.yamlfile, change all (YOUR_PROJECT) fields to reflect your current project ID. There should be 2 total.Make sure the changes are saved by going to File – Save from the Cloud Shell Editor menu.
Back in the command prompt, deploy your first deployment by typing the following:
gcloud deployment-manager deployments create single-vm --config vm-web.yamlOnce the deployment completes, go to the web console and verify that a Compute Engine instance has been created. You should be able to click the external IP address to view its website.
Still in the web console, delete your deployment by going to the top-left menu > Deployment Manager. Select your deployment and click DELETE.
- Deploy complex deployment with dependencies
We are now going to deploy a more complex deployment manager configuration that requires deploying resources in a certain order.
- Back in the Cloud Shell Editor, select the file
dependencies.yaml. - Under the resource for server-1, scroll down to the NetworkInterfaces properties and notice that the network and subnetwork have a reference field supplied. This tells Deployment Manager that deploying this instance depends on other resources in the same deployment being created first. You can view other dependencies throughout the rest of the config file.
- As before, change all the ‘(YOUR_PROJECT)’ fields to your current project ID. There are again two total, which you can find by the #CHANGE ME at the end of each affected line
- Make sure the changes are saved by going to File – Save from the Cloud Shell Editor menu.
Back in command prompt, deploy this configuration by typing the following:
gcloud deployment-manager deployments create vpcs --config dependencies.yaml
- Back in the Cloud Shell Editor, select the file
- View deployment manifest and delete deployment
After deployment is complete, let’s view the manifest:
Still in command prompt, view details of your deployment:
gcloud deployment-manager deployments describe vpcsLook for a field in the form of
manifest-(TIMESTAMP). Copy this field.Type in the following to view the full manifest, supplying the manifest ID you copied above:
gcloud deployment-manager manifests describe manifest-(TIMESTAMP) --deployment vpcsYou can also view the same manifests in the web console > Deployment Manager.
Delete your deployment using the following command:
gcloud deployment-manager deployments delete vpcs
- Create deployment with templates
In Cloud Shell, go to your templates directory:
cd templatesIn the Code Editor, view the
template-config.yamland the template.jinjafiles.This time, we do not need to change the project ID because the template files are able to automatically resolve the current project ID.
Deploy your template-enabled configuration by typing:
gcloud deployment-manager deployments create templates --config template-config.yamlView results in web console, and attempt to open website via external IP.
Delete deployment using one of the methods above.
