Creating a Basic Template

A basic configuration file might be enough for simple workloads but for more complex architectures or for configurations that you plan to reuse, you can break your configuration into templates.

A template is a separate file that is imported and used as a type in a configuration. You can use as many templates as you want in a configuration.

Templates allow you to separate your configuration out into different pieces that you can use and reuse across different deployments. Templates can be as generalized or specific as you need. With templates, you can also take advantage of features like template properties, environment variables, modules, and other template functionality to create dynamic configuration and template files.

For examples of templates that you can use in your own deployments, see the Deployment Manager GitHub repository.

Before you begin

Template syntax

Templates can be written in either Jinja 2.10.x or Python 3.x. Jinja maps more closely to the YAML syntax, so it might be easier to write templates in Jinja if you are more familiar with YAML.

You can also write template files in Python and take advantage of Python to programmatically generate parts of your templates. For example, you can use Python libraries to format template definitions. If you are familiar with Python, this might be a better format for you.

Deployment Manager accepts both Jinja and Python templates. You can import templates in both languages in the same configuration.

Creating a basic template

A template is a file you create, written in either Jinja or Python. For example, consider the following configuration file:

# Copyright 2016 Google Inc. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

resources:
- name: vm-created-by-deployment-manager
  type: compute.v1.instance
  properties:
    zone: us-central1-a
    machineType: zones/us-central1-a/machineTypes/n1-standard-1
    disks:
    - deviceName: boot
      type: PERSISTENT
      boot: true
      autoDelete: true
      initializeParams:
        sourceImage: projects/debian-cloud/global/images/family/debian-11
    networkInterfaces:
    - network: global/networks/default

This configuration is valid, but you can further simplify the configuration by breaking different parts as individual template files that you can reuse. To create a template based on the configuration above, pull out section for the resource in question and create a new Jinja or Python file.

The following snippets show the sections of templates that can simplify your deployment. For the full templates, click View on GitHub.

Jinja

- name: vm-template
  type: compute.v1.instance
  properties:
    zone: us-central1-a
    machineType: zones/us-central1-a/machineTypes/n1-standard-1
    disks:
    - deviceName: boot
      type: PERSISTENT
      boot: true
      autoDelete: true
      initializeParams:
        sourceImage: projects/debian-cloud/global/images/family/debian-11
    networkInterfaces:
    - network: global/networks/default

Python

Your Python templates must meet these requirements:

  • The template must define a method called GenerateConfig(context) or generate_config(context). If you use both method names in the same template, the generate_config() method takes precedence.

    The context object contains metadata about the deployment and your environment, such as the deployment's name, the current project, and so on. Learn more about using deployment-specific environment variables.

  • The method must return a Python dictionary.

Other than that, it is up to you to generate the contents of your template.

Example

resources.append({
    'name': 'vm-template',
    'type': 'compute.v1.instance',
    'properties': {
        'zone': 'us-central1-a',
        'machineType': 'zones/us-central1-a/machineTypes/n1-standard-1',
        'disks': [{
            'deviceName': 'boot',
            'type': 'PERSISTENT',
            'boot': True,
            'autoDelete': True,
            'initializeParams': {
                'sourceImage':
                    'projects/debian-cloud/global/images/family/debian-11'
            }
        }],
        'networkInterfaces': [{
            'network': 'global/networks/default'
        }]
    }
})

For the full Python example, see the Deployment Manager GitHub repository.

Importing the template

After creating a template, import it into your configuration file to use it. To import a template, add an imports section in your configuration, followed by the relative or absolute path from the current directory. For example, you can import your virtual machine template from the previous step by adding the following line to the top of your configuration:

imports:
  - path: path/to/my_vm_template.jinja

If you have a long file path, you can provide an optional name property as an alias for the file. You can use this name to reference the template later. If you do not provide the name, the template can be referenced using the path.

imports:
  - path: path/to/my_vm_template.jinja
    name: my_renamed_template.jinja

You can mix and match imports of multiple templates, regardless of whether the templates are Jinja or Python:

imports:
  - path: path/to/my_vm_template.jinja
    name: my_renamed_template.jinja
  - path: special_vm.py

If your template uses other templates as dependencies, import the dependent templates in your configuration as well:

imports:
  - path: path/to/my_vm_template.jinja
  - path: special_vm.py
  - path: base_vm.jinja

You can also import text files in order to inline the content. For example, if you create a file named resource_type.txt with the following string:

compute.v1.instance

Import it into your configuration and provide the content inline like so:

imports:
- path: resource_type.txt

resources:
- name: my-vm
  type: {{ imports["resource_type.txt"] }} # Resolves to "compute.v1.instance"
  properties:
    zone: us-central1-a
    machineType: zones/us-central1-a/machineTypes/f1-micro
    disks:
    - deviceName: boot
      type: PERSISTENT
      boot: true
      autoDelete: true
      initializeParams:
        sourceImage: projects/debian-cloud/global/images/family/debian-9
    networkInterfaces:
    - network: global/networks/default
      accessConfigs:
      - name: External NAT
        type: ONE_TO_ONE_NAT

Deploying templates

Once you import a template, use it as a type in your configuration:

Jinja

imports:
- path: vm-template.jinja

resources:
- name: my-vm
  type: vm-template.jinja

Python

imports:
- path: vm-template.py

resources:
- name: my-vm
  type: vm-template.py

If you did not provide a name for your template, call the template using the template path:

imports:
- path: path/to/my_vm_template.jinja

resources:
- name: my-first-virtual-machine
  type: path/to/my_vm_template.jinja

Deploying templates directly with the command-line tool

Instead of creating a top-level configuration file, Deployment Manager offers the ability to deploy a template directly in the Google Cloud CLI.

For example, the following request deploys a template called vm-template.jinja:

gcloud deployment-manager deployments create a-single-vm --template vm-template.jinja

If your template has template properties, you can also set these properties on the command-line using the --properties flag:

gcloud deployment-manager deployments create my-igm \
    --template vm-template.jinja \
    --properties zone:us-central1-a

Keep in mind that:

  • All values are parsed as YAML values. For example, version: 3 is passed in as an integer. If you want to specify it as a string, put escaped single quotes around the value, version: \'3\'.

  • Boolean values are case insensitive, so TRUE, true, and True are treated the same.

  • You must pass in all required properties defined by the template. You cannot provide just a subset of the properties. If certain properties have default values, you can omit the property from the command-line.

To specify multiple properties, provide comma-separated key:value pairs. It does not matter in what order you specify the pairs. For example:

gcloud deployment-manager deployments create my-igm 
--template vm-template.jinja
--properties zone:us-central1-a,machineType:n1-standard-1,image:debian-9

After running this command, Deployment Manager creates a deployment using the template you provided. You can confirm that the deployment has been created using the Google Cloud console or the gcloud CLI. For information on viewing a deployment, read Viewing a manifest.

What's next