This full example describes how to create a deployment with a virtual machine that
uses a container-optimized image. For more information on using containers with
Compute Engine, see
Container-Optimized Compute Engine Images
.
This walk-through describes how to:
- Create a simple container manifest.
- Create a configuration and template that uses a container image.
- Deploy your resources and verify that the deployment was successful.
Create a container manifest
To use containers, you must define a container manifest. The manifest describes
properties such as the container image, containers to launch, commands to
execute on boot, and ports to enable.
Create a file named
container_manifest.yaml
with the following contents:
This manifest creates a container named simple-echo that uses the
Hello Application
container image and
launches an echo server that listens on port 8080.
Create a template and configuration
Next, create a template that launches a virtual machine instance with a
container-optimized image. Create a file named
container_vm.[jinja|py]
with
the following contents:
Python
# 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.
"""Creates a Container VM with the provided Container manifest."""
COMPUTE_URL_BASE = 'https://www.googleapis.com/compute/v1/'
def GlobalComputeUrl(project, collection, name):
return ''.join([COMPUTE_URL_BASE, 'projects/', project,
'/global/', collection, '/', name])
def ZonalComputeUrl(project, zone, collection, name):
return ''.join([COMPUTE_URL_BASE, 'projects/', project,
'/zones/', zone, '/', collection, '/', name])
def GenerateConfig(context):
"""Generate configuration."""
res = []
base_name = (context.env['deployment'] + '-' +
context.env['name'])
# Properties for the container-based instance.
instance = {
'zone': context.properties['zone'],
'machineType': ZonalComputeUrl(context.env['project'],
context.properties['zone'],
'machineTypes',
'n1-standard-1'),
'metadata': {
'items': [{
'key': 'gce-container-declaration',
'value': context.imports[
context.properties['containerManifest']],
}]
},
'disks': [{
'deviceName': 'boot',
'type': 'PERSISTENT',
'autoDelete': True,
'boot': True,
'initializeParams': {
'diskName': base_name + '-disk',
'sourceImage': GlobalComputeUrl('cos-cloud',
'images',
context.properties[
'containerImage'])
},
}],
'networkInterfaces': [{
'accessConfigs': [{
'name': 'external-nat',
'type': 'ONE_TO_ONE_NAT'
}],
'network': GlobalComputeUrl(context.env['project'],
'networks',
'default')
}],
'serviceAccounts': [{
'email': 'default',
'scopes': [
"https://www.googleapis.com/auth/logging.write",
"https://www.googleapis.com/auth/monitoring.write"
]
}]
}
res.append({
'name': base_name,
'type': 'compute.v1.instance',
'properties': instance
})
# Resources to return.
resources = {
'resources': res,
}
return resources
Create the corresponding
schema file
,
which enforces the structure of the template:
Notice there are a number of parameters defined in this template including:
- The
deployment
,
name
, and
project
environment variables. Deployment
Manager automatically populates those variables without additional
action from you.
- The
zone
,
containerImage
, and
containerManifest
properties, which will be defined
in the configuration.
Create a configuration file named
container_vm.yaml
that
looks like the following:
Make sure to replace
ZONE_TO_RUN
with the desired zone for your virtual
machine. Notice that the file has also defined the container image to use and
the container manifest that you created earlier.
Deploy your virtual machine instance
Finally, deploy your virtual machine instance using the Google Cloud CLI:
gcloud deployment-manager deployments create my-container-deployment \
--config container_vm.yaml
Once the deployment has been created, you can view the details of your
deployment. For example:
$
gcloud deployment-manager deployments describe my-container-deployment
creationTimestamp: '2015-04-02T12:24:31.645-07:00'
fingerprint: ''
id: '8602410885927938432'
manifest: https://www.googleapis.com/deploymentmanager/v2/projects/myproject/global/deployments/my-container-deployment/manifests/manifest-1428002671645
name: my-container-deployment
state: DEPLOYED
resources:
NAME TYPE ID UPDATE_STATE ERRORS
my-container-deployment-my-container-vm compute.v1.instance 3775082612632070557 COMPLETED -
Verify that your instance is running
To test that your container instance started up, visit the virtual
machine's external IP address in your browser, which should print
hello world
:
Add a Compute Engine firewall rule to allow you to query traffic on the virtual
machine through port 8080:
gcloud compute firewall-rules create allow-8080 --allow tcp:8080
Get your instance's external IP address:
$
gcloud compute instances describe my-container-deployment-my-container-vm
...
name: my-container-vm-my-container-deployment
networkInterfaces:
- accessConfigs:
- kind: compute#accessConfig
name: external-nat
natIP: 104.197.8.138
type: ONE_TO_ONE_NAT
name: nic0
network: https://www.googleapis.com/compute/v1/projects/myproject/global/networks/default
networkIP: 10.240.97.220
scheduling:
automaticRestart: true
onHostMaintenance: MIGRATE
selfLink: https://www.googleapis.com/compute/v1/projects/myproject/zones/us-central1-a/instances/my-container-deployment-my-container-vm
status: RUNNING
tags:
fingerprint: 42WmSpB8rSM=
zone: https://www.googleapis.com/compute/v1/projects/myproject/zones/us-central1-a
...
In this case, the external IP is
104.197.8.138
.
In a browser window, enter in the external IP and port 8080 in the browser
bar. For example,
104.197.8.138:8080
.
If successful, you should see a
hello world
message.
(Optional) Delete your deployment
If you want to save on costs and no longer want or need your deployment, delete
your deployment.
gcloud deployment-manager deployments delete my-container-deployment
What's next
Explore more about Deployment Manager in the
Complete User Guides
or through the
API
.
Try out some other tutorials: