Usa los operadores de Google Kubernetes Engine

Cloud Composer 1 | Cloud Composer 2 | Cloud Composer 3

En esta página, se describe cómo usar los operadores de Google Kubernetes Engine para crear en Google Kubernetes Engine y para iniciar Pods de Kubernetes en esos clústeres.

Los operadores de Google Kubernetes Engine ejecutan pods de Kubernetes en un clúster específico, que puede ser un clúster separado que no está relacionado con tu entorno. En comparación, KubernetesPodOperator ejecuta pods de Kubernetes en el clúster de tu entorno.

En esta página, se muestra un ejemplo de DAG que crea un clúster de Google Kubernetes Engine con GKECreateClusterOperator, usa GKEStartPodOperator con las siguientes configuraciones y, luego, lo borra con GKEDeleteClusterOperator:

• Configuración mínima: establece solo los parámetros requeridos.
• Configuración de plantilla: utiliza parámetros que puedes agregar a una plantilla con Jinja.
• Configuración de afinidad de pods: Limita los nodos disponibles para programar pods.
• Configuración completa: incluye todas las configuraciones.

Antes de comenzar

Precaución: Si inicias Pods en el clúster de tu entorno, no uses Espacios de nombres de Cloud Composer (composer-*) distintos de composer-user-workloads. Estos espacios de nombres durante las operaciones de actualización del entorno. Esto podría provocar que los DAG deja de funcionar.

Configuración del operador de GKE

Para continuar con este ejemplo, coloca todo el archivo gke_operator.py en la carpeta dags/ de tu entorno o agrega el código relevante a un DAG.

Crea un clúster

Mediante el código que se muestra aquí, se crea un clúster de Google Kubernetes Engine con dos grupos de nodos, pool-0 y pool-1, cada uno de los cuales tiene un nodo. Si es necesario, puedes configurar otros parámetros de la API de Google Kubernetes Engine como parte de body.

Nota: Te recomendamos que uses clústeres regionales en Airflow 2. Los clústeres zonales son están más expuestos a fallas zonales. Por ejemplo, puedes usar us-central1. como una zona para el clúster en lugar de us-central1-a.

Antes del lanzamiento de la versión 5.1.0 de apache-airflow-providers-google, no fue posible pasar el objeto node_pools en el GKECreateClusterOperator. Si usas Airflow 2, asegúrate de que tu entorno debe usar la versión 5.1.0 de apache-airflow-providers-google o una posterior. Puedes instalar una versión más reciente de este paquete de PyPI si especificas apache-airflow-providers-google y >=5.1.0 como la versión requerida.

# TODO(developer): update with your values
PROJECT_ID = "my-project-id"
# It is recommended to use regional clusters for increased reliability
# though passing a zone in the location parameter is also valid
CLUSTER_REGION = "us-west1"
CLUSTER_NAME = "example-cluster"
CLUSTER = {
    "name": CLUSTER_NAME,
    "node_pools": [
        {"name": "pool-0", "initial_node_count": 1},
        {"name": "pool-1", "initial_node_count": 1},
    ],
}
create_cluster = GKECreateClusterOperator(
    task_id="create_cluster",
    project_id=PROJECT_ID,
    location=CLUSTER_REGION,
    body=CLUSTER,
)

Inicia cargas de trabajo en el clúster

Las siguientes secciones explican cada configuración de GKEStartPodOperator en el ejemplo. Para obtener información sobre cada variable de configuración, consulta la referencia de Airflow para los operadores de GKE.

from airflow import models
from airflow.providers.google.cloud.operators.kubernetes_engine import (
    GKECreateClusterOperator,
    GKEDeleteClusterOperator,
    GKEStartPodOperator,
)
from airflow.utils.dates import days_ago

from kubernetes.client import models as k8s_models


with models.DAG(
    "example_gcp_gke",
    schedule_interval=None,  # Override to match your needs
    start_date=days_ago(1),
    tags=["example"],
) as dag:
    # TODO(developer): update with your values
    PROJECT_ID = "my-project-id"
    # It is recommended to use regional clusters for increased reliability
    # though passing a zone in the location parameter is also valid
    CLUSTER_REGION = "us-west1"
    CLUSTER_NAME = "example-cluster"
    CLUSTER = {
        "name": CLUSTER_NAME,
        "node_pools": [
            {"name": "pool-0", "initial_node_count": 1},
            {"name": "pool-1", "initial_node_count": 1},
        ],
    }
    create_cluster = GKECreateClusterOperator(
        task_id="create_cluster",
        project_id=PROJECT_ID,
        location=CLUSTER_REGION,
        body=CLUSTER,
    )

    kubernetes_min_pod = GKEStartPodOperator(
        # The ID specified for the task.
        task_id="pod-ex-minimum",
        # Name of task you want to run, used to generate Pod ID.
        name="pod-ex-minimum",
        project_id=PROJECT_ID,
        location=CLUSTER_REGION,
        cluster_name=CLUSTER_NAME,
        # Entrypoint of the container, if not specified the Docker container's
        # entrypoint is used. The cmds parameter is templated.
        cmds=["echo"],
        # The namespace to run within Kubernetes, default namespace is
        # `default`.
        namespace="default",
        # Docker image specified. Defaults to hub.docker.com, but any fully
        # qualified URLs will point to a custom repository. Supports private
        # gcr.io images if the Composer Environment is under the same
        # project-id as the gcr.io images and the service account that Composer
        # uses has permission to access the Google Container Registry
        # (the default service account has permission)
        image="gcr.io/gcp-runtimes/ubuntu_18_0_4",
    )

    kubenetes_template_ex = GKEStartPodOperator(
        task_id="ex-kube-templates",
        name="ex-kube-templates",
        project_id=PROJECT_ID,
        location=CLUSTER_REGION,
        cluster_name=CLUSTER_NAME,
        namespace="default",
        image="bash",
        # All parameters below are able to be templated with jinja -- cmds,
        # arguments, env_vars, and config_file. For more information visit:
        # https://airflow.apache.org/docs/apache-airflow/stable/macros-ref.html
        # Entrypoint of the container, if not specified the Docker container's
        # entrypoint is used. The cmds parameter is templated.
        cmds=["echo"],
        # DS in jinja is the execution date as YYYY-MM-DD, this docker image
        # will echo the execution date. Arguments to the entrypoint. The docker
        # image's CMD is used if this is not provided. The arguments parameter
        # is templated.
        arguments=["{{ ds }}"],
        # The var template variable allows you to access variables defined in
        # Airflow UI. In this case we are getting the value of my_value and
        # setting the environment variable `MY_VALUE`. The pod will fail if
        # `my_value` is not set in the Airflow UI.
        env_vars={"MY_VALUE": "{{ var.value.my_value }}"},
    )

    kubernetes_affinity_ex = GKEStartPodOperator(
        task_id="ex-pod-affinity",
        project_id=PROJECT_ID,
        location=CLUSTER_REGION,
        cluster_name=CLUSTER_NAME,
        name="ex-pod-affinity",
        namespace="default",
        image="perl",
        cmds=["perl"],
        arguments=["-Mbignum=bpi", "-wle", "print bpi(2000)"],
        # affinity allows you to constrain which nodes your pod is eligible to
        # be scheduled on, based on labels on the node. In this case, if the
        # label 'cloud.google.com/gke-nodepool' with value
        # 'nodepool-label-value' or 'nodepool-label-value2' is not found on any
        # nodes, it will fail to schedule.
        affinity={
            "nodeAffinity": {
                # requiredDuringSchedulingIgnoredDuringExecution means in order
                # for a pod to be scheduled on a node, the node must have the
                # specified labels. However, if labels on a node change at
                # runtime such that the affinity rules on a pod are no longer
                # met, the pod will still continue to run on the node.
                "requiredDuringSchedulingIgnoredDuringExecution": {
                    "nodeSelectorTerms": [
                        {
                            "matchExpressions": [
                                {
                                    # When nodepools are created in Google Kubernetes
                                    # Engine, the nodes inside of that nodepool are
                                    # automatically assigned the label
                                    # 'cloud.google.com/gke-nodepool' with the value of
                                    # the nodepool's name.
                                    "key": "cloud.google.com/gke-nodepool",
                                    "operator": "In",
                                    # The label key's value that pods can be scheduled
                                    # on.
                                    "values": [
                                        "pool-1",
                                    ],
                                }
                            ]
                        }
                    ]
                }
            }
        },
    )
    kubernetes_full_pod = GKEStartPodOperator(
        task_id="ex-all-configs",
        name="full",
        project_id=PROJECT_ID,
        location=CLUSTER_REGION,
        cluster_name=CLUSTER_NAME,
        namespace="default",
        image="perl:5.34.0",
        # Entrypoint of the container, if not specified the Docker container's
        # entrypoint is used. The cmds parameter is templated.
        cmds=["perl"],
        # Arguments to the entrypoint. The docker image's CMD is used if this
        # is not provided. The arguments parameter is templated.
        arguments=["-Mbignum=bpi", "-wle", "print bpi(2000)"],
        # The secrets to pass to Pod, the Pod will fail to create if the
        # secrets you specify in a Secret object do not exist in Kubernetes.
        secrets=[],
        # Labels to apply to the Pod.
        labels={"pod-label": "label-name"},
        # Timeout to start up the Pod, default is 120.
        startup_timeout_seconds=120,
        # The environment variables to be initialized in the container
        # env_vars are templated.
        env_vars={"EXAMPLE_VAR": "/example/value"},
        # If true, logs stdout output of container. Defaults to True.
        get_logs=True,
        # Determines when to pull a fresh image, if 'IfNotPresent' will cause
        # the Kubelet to skip pulling an image if it already exists. If you
        # want to always pull a new image, set it to 'Always'.
        image_pull_policy="Always",
        # Annotations are non-identifying metadata you can attach to the Pod.
        # Can be a large range of data, and can include characters that are not
        # permitted by labels.
        annotations={"key1": "value1"},
        # Optional resource specifications for Pod, this will allow you to
        # set both cpu and memory limits and requirements.
        # Prior to Airflow 2.3 and the cncf providers package 5.0.0
        # resources were passed as a dictionary. This change was made in
        # https://github.com/apache/airflow/pull/27197
        # Additionally, "memory" and "cpu" were previously named
        # "limit_memory" and "limit_cpu"
        # resources={'limit_memory': "250M", 'limit_cpu': "100m"},
        container_resources=k8s_models.V1ResourceRequirements(
            limits={"memory": "250M", "cpu": "100m"},
        ),
        # If true, the content of /airflow/xcom/return.json from container will
        # also be pushed to an XCom when the container ends.
        do_xcom_push=False,
        # List of Volume objects to pass to the Pod.
        volumes=[],
        # List of VolumeMount objects to pass to the Pod.
        volume_mounts=[],
        # Affinity determines which nodes the Pod can run on based on the
        # config. For more information see:
        # https://kubernetes.io/docs/concepts/configuration/assign-pod-node/
        affinity={},
    )
    delete_cluster = GKEDeleteClusterOperator(
        task_id="delete_cluster",
        name=CLUSTER_NAME,
        project_id=PROJECT_ID,
        location=CLUSTER_REGION,
    )

    create_cluster >> kubernetes_min_pod >> delete_cluster
    create_cluster >> kubernetes_full_pod >> delete_cluster
    create_cluster >> kubernetes_affinity_ex >> delete_cluster
    create_cluster >> kubenetes_template_ex >> delete_cluster

Configuración mínima

Para iniciar un pod en tu clúster de GKE con GKEStartPodOperator, solo las opciones project_id, location, cluster_name, name, namespace, image y task_id son obligatorias.

Cuando colocas el siguiente fragmento de código en un DAG, la tarea pod-ex-minimum funciona siempre que los parámetros enumerados antes estén definidos y sean válidos.

# TODO(developer): update with your values
PROJECT_ID = "my-project-id"
# It is recommended to use regional clusters for increased reliability
# though passing a zone in the location parameter is also valid
CLUSTER_REGION = "us-west1"
CLUSTER_NAME = "example-cluster"
kubernetes_min_pod = GKEStartPodOperator(
    # The ID specified for the task.
    task_id="pod-ex-minimum",
    # Name of task you want to run, used to generate Pod ID.
    name="pod-ex-minimum",
    project_id=PROJECT_ID,
    location=CLUSTER_REGION,
    cluster_name=CLUSTER_NAME,
    # Entrypoint of the container, if not specified the Docker container's
    # entrypoint is used. The cmds parameter is templated.
    cmds=["echo"],
    # The namespace to run within Kubernetes, default namespace is
    # `default`.
    namespace="default",
    # Docker image specified. Defaults to hub.docker.com, but any fully
    # qualified URLs will point to a custom repository. Supports private
    # gcr.io images if the Composer Environment is under the same
    # project-id as the gcr.io images and the service account that Composer
    # uses has permission to access the Google Container Registry
    # (the default service account has permission)
    image="gcr.io/gcp-runtimes/ubuntu_18_0_4",
)

Configuración de la plantilla

Airflow admite el uso de plantillas de Jinja. Debes declarar las variables obligatorias (task_id, name, namespace y image) con el operador. Como se muestra en el siguiente ejemplo, puedes crear plantillas de todos los demás parámetros con Jinja, incluidos cmds, arguments y env_vars.

Si no cambias el DAG ni tu entorno, la tarea ex-kube-templates falla. Configura una variable de Airflow llamada my_value para que este DAG se realice con éxito.

Para configurar my_value con gcloud o la IU de Airflow:

gcloud

Para Airflow 2, ingresa el siguiente comando:

gcloud composer environments run ENVIRONMENT \
    --location LOCATION \
    variables set -- \
    my_value example_value

Reemplaza lo siguiente:

• ENVIRONMENT por el nombre del entorno.
• LOCATION por la región en la que se encuentra el entorno

IU de Airflow

En la IU de Airflow 2, haz lo siguiente:

1. En la barra de herramientas, selecciona Administrador > Variables.

2. En la página Variable de lista, haz clic en Agregar un registro nuevo.

3. En la página Agregar variable, ingresa la siguiente información:

   • Key: my_value
   • Val: example_value

4. Haz clic en Guardar.

Configuración de plantilla:

# TODO(developer): update with your values
PROJECT_ID = "my-project-id"
# It is recommended to use regional clusters for increased reliability
# though passing a zone in the location parameter is also valid
CLUSTER_REGION = "us-west1"
CLUSTER_NAME = "example-cluster"
kubenetes_template_ex = GKEStartPodOperator(
    task_id="ex-kube-templates",
    name="ex-kube-templates",
    project_id=PROJECT_ID,
    location=CLUSTER_REGION,
    cluster_name=CLUSTER_NAME,
    namespace="default",
    image="bash",
    # All parameters below are able to be templated with jinja -- cmds,
    # arguments, env_vars, and config_file. For more information visit:
    # https://airflow.apache.org/docs/apache-airflow/stable/macros-ref.html
    # Entrypoint of the container, if not specified the Docker container's
    # entrypoint is used. The cmds parameter is templated.
    cmds=["echo"],
    # DS in jinja is the execution date as YYYY-MM-DD, this docker image
    # will echo the execution date. Arguments to the entrypoint. The docker
    # image's CMD is used if this is not provided. The arguments parameter
    # is templated.
    arguments=["{{ ds }}"],
    # The var template variable allows you to access variables defined in
    # Airflow UI. In this case we are getting the value of my_value and
    # setting the environment variable `MY_VALUE`. The pod will fail if
    # `my_value` is not set in the Airflow UI.
    env_vars={"MY_VALUE": "{{ var.value.my_value }}"},
)

Configuración de afinidad de los pods

Si configuras el parámetro affinity en el GKEStartPodOperator, puedes controlar en qué nodos se programan los pods (por ejemplo, puedes especificar un grupo de nodos en particular). Cuando creaste el clúster, creaste dos grupos de nodos llamados pool-0 y pool-1. Este operador determina que los pods deben ejecutarse solo en pool-1.

Ubicación del inicio del pod de Kubernetes de Cloud Composer con afinidad de pods (haz clic para ampliar)

# TODO(developer): update with your values
PROJECT_ID = "my-project-id"
# It is recommended to use regional clusters for increased reliability
# though passing a zone in the location parameter is also valid
CLUSTER_REGION = "us-west1"
CLUSTER_NAME = "example-cluster"
kubernetes_affinity_ex = GKEStartPodOperator(
    task_id="ex-pod-affinity",
    project_id=PROJECT_ID,
    location=CLUSTER_REGION,
    cluster_name=CLUSTER_NAME,
    name="ex-pod-affinity",
    namespace="default",
    image="perl",
    cmds=["perl"],
    arguments=["-Mbignum=bpi", "-wle", "print bpi(2000)"],
    # affinity allows you to constrain which nodes your pod is eligible to
    # be scheduled on, based on labels on the node. In this case, if the
    # label 'cloud.google.com/gke-nodepool' with value
    # 'nodepool-label-value' or 'nodepool-label-value2' is not found on any
    # nodes, it will fail to schedule.
    affinity={
        "nodeAffinity": {
            # requiredDuringSchedulingIgnoredDuringExecution means in order
            # for a pod to be scheduled on a node, the node must have the
            # specified labels. However, if labels on a node change at
            # runtime such that the affinity rules on a pod are no longer
            # met, the pod will still continue to run on the node.
            "requiredDuringSchedulingIgnoredDuringExecution": {
                "nodeSelectorTerms": [
                    {
                        "matchExpressions": [
                            {
                                # When nodepools are created in Google Kubernetes
                                # Engine, the nodes inside of that nodepool are
                                # automatically assigned the label
                                # 'cloud.google.com/gke-nodepool' with the value of
                                # the nodepool's name.
                                "key": "cloud.google.com/gke-nodepool",
                                "operator": "In",
                                # The label key's value that pods can be scheduled
                                # on.
                                "values": [
                                    "pool-1",
                                ],
                            }
                        ]
                    }
                ]
            }
        }
    },
)

Configuración completa

En este ejemplo, se muestran todas las variables que puedes configurar en GKEStartPodOperator. No es necesario modificar el código para que la tarea ex-all-configs se realice con éxito.

Para obtener detalles sobre cada variable, consulta la referencia de Airflow para los operadores de GKE.

# TODO(developer): update with your values
PROJECT_ID = "my-project-id"
# It is recommended to use regional clusters for increased reliability
# though passing a zone in the location parameter is also valid
CLUSTER_REGION = "us-west1"
CLUSTER_NAME = "example-cluster"
kubernetes_full_pod = GKEStartPodOperator(
    task_id="ex-all-configs",
    name="full",
    project_id=PROJECT_ID,
    location=CLUSTER_REGION,
    cluster_name=CLUSTER_NAME,
    namespace="default",
    image="perl:5.34.0",
    # Entrypoint of the container, if not specified the Docker container's
    # entrypoint is used. The cmds parameter is templated.
    cmds=["perl"],
    # Arguments to the entrypoint. The docker image's CMD is used if this
    # is not provided. The arguments parameter is templated.
    arguments=["-Mbignum=bpi", "-wle", "print bpi(2000)"],
    # The secrets to pass to Pod, the Pod will fail to create if the
    # secrets you specify in a Secret object do not exist in Kubernetes.
    secrets=[],
    # Labels to apply to the Pod.
    labels={"pod-label": "label-name"},
    # Timeout to start up the Pod, default is 120.
    startup_timeout_seconds=120,
    # The environment variables to be initialized in the container
    # env_vars are templated.
    env_vars={"EXAMPLE_VAR": "/example/value"},
    # If true, logs stdout output of container. Defaults to True.
    get_logs=True,
    # Determines when to pull a fresh image, if 'IfNotPresent' will cause
    # the Kubelet to skip pulling an image if it already exists. If you
    # want to always pull a new image, set it to 'Always'.
    image_pull_policy="Always",
    # Annotations are non-identifying metadata you can attach to the Pod.
    # Can be a large range of data, and can include characters that are not
    # permitted by labels.
    annotations={"key1": "value1"},
    # Optional resource specifications for Pod, this will allow you to
    # set both cpu and memory limits and requirements.
    # Prior to Airflow 2.3 and the cncf providers package 5.0.0
    # resources were passed as a dictionary. This change was made in
    # https://github.com/apache/airflow/pull/27197
    # Additionally, "memory" and "cpu" were previously named
    # "limit_memory" and "limit_cpu"
    # resources={'limit_memory': "250M", 'limit_cpu': "100m"},
    container_resources=k8s_models.V1ResourceRequirements(
        limits={"memory": "250M", "cpu": "100m"},
    ),
    # If true, the content of /airflow/xcom/return.json from container will
    # also be pushed to an XCom when the container ends.
    do_xcom_push=False,
    # List of Volume objects to pass to the Pod.
    volumes=[],
    # List of VolumeMount objects to pass to the Pod.
    volume_mounts=[],
    # Affinity determines which nodes the Pod can run on based on the
    # config. For more information see:
    # https://kubernetes.io/docs/concepts/configuration/assign-pod-node/
    affinity={},
)

Borra el clúster

El código que se muestra aquí borra el clúster que se creó al comienzo de la guía.

delete_cluster = GKEDeleteClusterOperator(
    task_id="delete_cluster",
    name=CLUSTER_NAME,
    project_id=PROJECT_ID,
    location=CLUSTER_REGION,
)

¿Qué sigue?

• Usa KubernetesPodOperator
• Administra DAG