Crea una implementación optimizada para contenedores

En este ejemplo completo se describe cómo crear una implementación con una máquina virtual que utiliza una imagen optimizada para el contenedor. Para obtener más información sobre cómo usar contenedores con Compute Engine, consulta Imágenes de Compute Engine optimizadas para el contenedor.

En esta guía se describe cómo:

  1. Crear un manifiesto del contenedor simple
  2. Crear una configuración y una plantilla que utilicen un imagen de contenedor
  3. Implementar los recursos y verificar que la implementación se haya realizado correctamente

Crea un manifiesto del contenedor

Para utilizar contenedores, debes definir un manifiesto del contenedor. El manifiesto describe propiedades como la imagen de contenedor, los contenedores que se iniciarán, los comandos que se ejecutarán en el inicio y los puertos que se habilitarán.

Crea un archivo llamado container_manifest.yaml con el siguiente contenido:

# 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.

# This is a container manifest, as described here:
#   https://cloud.google.com/compute/docs/containers/container_vms
apiVersion: v1
kind: Pod
metadata:
  name: simple-echo
spec:
  containers:
    - name: simple-echo
      image: gcr.io/google-samples/hello-app:2.0
      imagePullPolicy: Always
      ports:
        - containerPort: 8080
          hostPort: 8080

Este manifiesto crea un contenedor llamado simple-echo que usa la imagen del contenedor Hello Application y ejecuta un servidor Echo que se escucha en el puerto 8080.

Crea una plantilla y una configuración

A continuación, crea una plantilla que inicie una instancia de máquina virtual con una imagen optimizada para el contenedor: Crea un archivo llamado container_vm.[jinja|py] con los siguientes contenidos:

Jinja



{% set COMPUTE_URL_BASE = 'https://www.googleapis.com/compute/v1/' %}
{% set BASE_NAME = env['deployment'] + '-' + env['name'] %}

{% macro GlobalComputeUrl(project, collection, name) -%}
{{ COMPUTE_URL_BASE }}projects/{{ project }}/global/{{ collection }}/{{ name }}
{%- endmacro %}

{% macro ZonalComputeUrl(project, zone, collection, name) -%}
{{ COMPUTE_URL_BASE }}projects/{{ project }}/zones/{{ zone }}/{{ collection }}/{{ name }}
{%- endmacro %}

resources:
- name: {{ BASE_NAME }}
  type: compute.v1.instance
  properties:
    zone: {{ properties['zone'] }}
    machineType: {{ ZonalComputeUrl(env['project'], properties['zone'], 'machineTypes', 'n1-standard-1') }}
    metadata:
      items:
        - key: gce-container-declaration
          value: |
            {{ imports[properties['containerManifest']]|indent(12) }}
    disks:
      - deviceName: boot
        type: PERSISTENT
        autoDelete: true
        boot: true
        initializeParams:
          diskName: {{ BASE_NAME }}-disk
          sourceImage: {{ GlobalComputeUrl('cos-cloud', 'images', properties['containerImage']) }}
    networkInterfaces:
      - accessConfigs:
          - name: external-nat
            type: ONE_TO_ONE_NAT
        network: {{ GlobalComputeUrl(env['project'],  'networks', 'default') }}
    serviceAccounts:
      - email: default
        scopes:
        - https://www.googleapis.com/auth/logging.write
        - https://www.googleapis.com/auth/monitoring.write

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

Crea el archivo de esquema correspondiente, que aplique la estructura de la plantilla:

Jinja

# 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.

info:
  title: Container VM
  author: Google Inc.
  description: Creates a Container VM with the provided Container manifest.

required:
  - zone
  - containerImage
  - containerManifest

properties:
  zone:
    description: Zone in which this VM will run
    type: string
  containerImage:
    description: Name of the Google Cloud Container VM Image
    type: string
  containerManifest:
    description: String containing the Container Manifest in YAML
    type: string

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.

info:
  title: Container VM
  author: Google Inc.
  description: Creates a Container VM with the provided Container manifest.

required:
  - zone
  - containerImage
  - containerManifest

properties:
  zone:
    description: Zone in which this VM will run
    type: string
  containerImage:
    description: Name of the Google Cloud Container VM Image
    type: string
  containerManifest:
    description: String containing the Container Manifest in YAML
    type: string

Ten en cuenta que hay varios parámetros definidos en esta plantilla, que incluyen los siguientes:

  • Las variables de entorno deployment, name y project. Deployment Manager propaga automáticamente esas variables sin ninguna intervención adicional de tu parte.
  • Las propiedades zone, containerImage y containerManifest, que se definirán en la configuración.

Crea un archivo de configuración llamado container_vm.yaml que se vea así:

Jinja

# 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.

imports:
- path: ../../common/container_manifest.yaml
  name: container_manifest
- path: container_vm.jinja

resources:
  - name: my-container-vm
    type: container_vm.jinja
    properties:
      zone: ZONE_TO_RUN
      containerImage: family/cos-stable
      containerManifest: container_manifest

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.

imports:
- path: ../../common/container_manifest.yaml
  name: container_manifest
- path: container_vm.py

resources:
  - name: my-container-vm
    type: container_vm.py
    properties:
      zone: ZONE_TO_RUN
      containerImage: family/cos-stable
      containerManifest: container_manifest

Asegúrate de reemplazar ZONE_TO_RUN por la zona deseada para la máquina virtual. Ten en cuenta que en el archivo también se definió la imagen de contenedor que se utilizará y el manifiesto del contenedor que creaste con anterioridad.

Implementa la instancia de máquina virtual

Por último, implementa tu instancia de máquina virtual usando Google Cloud CLI:

gcloud deployment-manager deployments create my-container-deployment \
  --config container_vm.yaml

Una vez que se haya creado la implementación, podrás ver los detalles de la implementación. Por ejemplo:

$ 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     -

Verifica que la instancia se esté ejecutando

Para probar que tu instancia de contenedor se inició, visita la dirección IP externa de la máquina virtual en tu navegador, en la que se debería mostrar hello world:

  1. Agrega una regla de firewall de Compute Engine para que te permita consultar el tráfico en la máquina virtual a través del puerto 8080:

    gcloud compute firewall-rules create allow-8080 --allow tcp:8080
    
  2. Obtén la dirección IP externa de la instancia:

    $ 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
    ...

    En este caso, la IP externa es 104.197.8.138.

  3. En la ventana del navegador, ingresa la IP externa y el puerto 8080 en la barra del navegador. Por ejemplo, 104.197.8.138:8080.

    Si se ejecuta con éxito, deberías ver un mensaje hello world.

(Opcional) Borra la implementación

Si deseas reducir costos y ya no quieres o no necesitas la implementación, bórrala.

gcloud deployment-manager deployments delete my-container-deployment

Pasos siguientes

Obtén más información sobre Deployment Manager en las Guías del usuario completas o a través de la API.

Prueba algunos otros instructivos: