Creazione di un deployment ottimizzato per i container

Questo esempio completo descrive come creare un deployment con una macchina virtuale che utilizza un'immagine ottimizzata per i container. Per ulteriori informazioni sull'utilizzo dei container con Compute Engine, consulta Immagini Compute Engine ottimizzate per i container.

Questa procedura dettagliata descrive come:

1. Creare un manifest del container semplice.
2. Creare una configurazione e un modello che utilizzano un'immagine container.
3. Esegui il deployment delle risorse e verifica che il deployment sia riuscito.

crea un manifest del container

Per utilizzare i container, devi definire un manifest del container. Il manifest descrive proprietà come l'immagine container, i container da avviare, i comandi da eseguire all'avvio e le porte da abilitare.

Crea un file denominato container_manifest.yaml con il seguente contenuto:

# 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

Questo manifest crea un container denominato simple-echo che utilizza l'immagine container Hello Application e avvia un server echo in ascolto sulla porta 8080.

Creare un modello e una configurazione

Quindi, crea un modello che avvia un'istanza di macchina virtuale con un'immagine ottimizzata per i container. Crea un file denominato container_vm.[jinja|py] con il seguente contenuto:

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 il file di schema corrispondente, che applica la struttura del modello:

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

Tieni presente che in questo modello sono definiti diversi parametri, tra cui:

• Le variabili di ambiente deployment, name e project. Deployment Manager compila automaticamente queste variabili senza ulteriori azioni da parte tua.
• Le proprietà zone, containerImage e containerManifest, che verranno definite nella configurazione.

Crea un file di configurazione denominato container_vm.yaml, simile al seguente:

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

Assicurati di sostituire ZONE_TO_RUN con la zona desiderata per la tua macchina virtuale. Nota che il file ha anche definito l'immagine container da utilizzare e il manifest del container creato in precedenza.

Esegui il deployment dell'istanza di una macchina virtuale

Infine, esegui il deployment dell'istanza della macchina virtuale utilizzando Google Cloud CLI:

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

Una volta creato il deployment, puoi visualizzare i relativi dettagli. Ad esempio:

$ 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 che l'istanza sia in esecuzione

Per verificare che la tua istanza di container sia stata avviata, visita l'indirizzo IP esterno della macchina virtuale nel tuo browser, che dovrebbe stampare hello world:

1. Aggiungi una regola firewall di Compute Engine per consentire di eseguire query sul traffico sulla macchina virtuale attraverso la porta 8080:

   gcloud compute firewall-rules create allow-8080 --allow tcp:8080
   

2. Recupera l'indirizzo IP esterno della tua istanza:

   $ 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 questo caso, l'IP esterno è 104.197.8.138.

3. In una finestra del browser, inserisci l'IP esterno e la porta 8080 nella barra del browser. Ad esempio, 104.197.8.138:8080.

   Se l'operazione va a buon fine, dovresti visualizzare un messaggio hello world.

(Facoltativo) Eliminare il deployment

Se vuoi risparmiare sui costi ma non vuoi più o non hai più bisogno del deployment, eliminalo.

gcloud deployment-manager deployments delete my-container-deployment

Passaggi successivi

Scopri di più su Deployment Manager nelle guide per l'utente complete o tramite l'API.

Prova altri tutorial:

• Crea un'applicazione logbook con bilanciamento del carico di rete.
• Crea un'applicazione logbook con bilanciamento del carico HTTP.