El producto descrito en esta documentación, Anthos Clusters en AWS (generación anterior), se encuentra actualmente en modo de mantenimiento . Todas las nuevas instalaciones deben usar el producto de la generación actual, Anthos Clusters en AWS .
Para conectarse a sus recursos de GKE en AWS, siga estos pasos. Seleccione si ya tiene una VPC de AWS (o una conexión directa a su VPC) o si creó una VPC dedicada al crear su servicio de administración.
VPC existente
Si tiene una conexión directa o VPN a una VPC existente, omita la línea env HTTP_PROXY=http://localhost:8118 de los comandos de este tema.
VPC dedicada
Cuando crea un servicio de administración en una VPC dedicada, GKE en AWS incluye un host bastión en una subred pública.
Para conectarse a su servicio de administración, realice los siguientes pasos:
Para abrir el túnel, ejecute el script bastion-tunnel.sh . El túnel redirecciona a localhost:8118 .
Para abrir un túnel al host bastión, ejecute el siguiente comando:
./bastion-tunnel.sh-N
Los mensajes del túnel SSH aparecen en esta ventana. Cuando esté listo para cerrar la conexión, detenga el proceso usando Control+C o cerrando la ventana.
Abra una nueva terminal y cámbiese a su directorio anthos-aws .
cd anthos-aws
Comprueba que puedes conectarte al clúster con kubectl .
La salida incluye la URL del servidor API del servicio de administración.
Eliminar clústeres de usuarios
Se usa kubectl para eliminar clústeres de usuarios. GKE en AWS cierra correctamente todos los nodos. Los pods reciben un SIGTERM y, tras un periodo de gracia, un SIGKILL . Una vez que los pods se detienen, el servicio de administración comienza a eliminar recursos de AWS.
Para eliminar un clúster de usuarios, realice los siguientes pasos:
Desde su directorio anthos-aws , use anthos-gke para cambiar el contexto a su clúster de usuarios.
cd anthos-aws
env HTTPS_PROXY=http://localhost:8118 \
anthos-gke aws clusters get-credentials CLUSTER_NAME
Reemplace CLUSTER_NAME con el nombre de su clúster de usuarios.
Utilice kubectl delete para eliminar cualquier servicio o PersistentVolumeClaim restante que haya creado. Por ejemplo, para eliminar el LoadBalancer my-lb-service , ejecute el siguiente comando:
Al crear un servicio de administración, anthos-gke crea un bucket de S3 en AWS con archivos binarios para la creación de clústeres. Este bucket se reutiliza al crear nuevos clústeres de usuarios. Si ya no usa GKE en AWS, puede eliminar el bucket después de eliminar el servicio de administración.
Para eliminar el bucket de GKE en AWS S3, realice los siguientes pasos:
Para encontrar el nombre del depósito, busque el valor de gcs_download_bucket en terraform-workspace/terraform.tfvars.json con grep .
grepagent_s3_bucketterraform.tfvars.json
Elimine este bucket s3 y su contenido con el comando aws , reemplazando bucket-name con el nombre de su bucket del comando anterior.
[[["Es fácil de entender","easyToUnderstand","thumb-up"],["Me ofreció una solución al problema","solvedMyProblem","thumb-up"],["Otro","otherUp","thumb-up"]],[["Es difícil de entender","hardToUnderstand","thumb-down"],["La información o el código de muestra no son correctos","incorrectInformationOrSampleCode","thumb-down"],["Me faltan las muestras o la información que necesito","missingTheInformationSamplesINeed","thumb-down"],["Problema de traducción","translationIssue","thumb-down"],["Otro","otherDown","thumb-down"]],["Última actualización: 2025-06-12 (UTC)."],[],[],null,["# Uninstalling GKE on AWS\n\nThis topic describes how to uninstall GKE on AWS.\n\nBefore you begin\n----------------\n\nTo perform these steps,\n[authenticate to your GKE on AWS management service](/kubernetes-engine/multi-cloud/docs/aws/previous-generation/how-to/installing-management#connect).\n\nTo connect to your GKE on AWS resources, perform the following\nsteps. Select if you have an existing AWS VPC (or direct connection to\nyour VPC) or created a dedicated VPC when creating your management service. \n\n### Existing VPC\n\nIf you have a direct or VPN connection to an existing VPC, omit the line\n`env HTTP_PROXY=http://localhost:8118` from commands in this topic.\n\n### Dedicated VPC\n\nWhen you create a management service in a dedicated VPC,\nGKE on AWS includes a\n[bastion](https://en.wikipedia.org/wiki/Bastion_host) host in a\npublic subnet.\n| **Important:** If you restart your terminal session or the SSH connection is lost, you need to re-launch the `bastion-tunnel.sh` script.\n\nTo connect to your management service, perform the following steps:\n\n1. Change to the directory with your GKE on AWS configuration.\n You created this directory when\n [Installing the management service](/kubernetes-engine/multi-cloud/docs/aws/previous-generation/how-to/installing-management).\n\n ```sh\n cd anthos-aws\n ```\n\n \u003cbr /\u003e\n\n2. To open the tunnel, run the `bastion-tunnel.sh` script. The tunnel forwards\n to `localhost:8118`.\n\n To open a tunnel to the bastion host, run the following command: \n\n ./bastion-tunnel.sh -N\n\n Messages from the SSH tunnel appear in this window. When you are ready to\n close the connection, stop the process by using \u003ckbd\u003eControl+C\u003c/kbd\u003e or\n closing the window.\n3. Open a new terminal and change into your `anthos-aws` directory.\n\n ```sh\n cd anthos-aws\n ```\n4. Check that you're able to connect to the cluster with `kubectl`.\n\n env HTTPS_PROXY=http://localhost:8118 \\\n kubectl cluster-info\n\n The output includes the URL for the management service API server.\n\nDeleting user clusters\n----------------------\n\nYou use `kubectl` to delete user clusters. GKE on AWS gracefully\nshuts down all nodes. Pods receive a `SIGTERM` and then, after a grace period, a\n`SIGKILL`. After the pods stop, the management service starts removing AWS\nresources.\n| **Important:** If you do not delete all your user clusters and their associated resources, you cannot use `anthos-gke` to [remove the management service](#removing_a_management_service).\n\nTo delete a user cluster, perform the following steps:\n\n1. From your `anthos-aws` directory, use\n `anthos-gke` to switch context to your user cluster.\n\n ```sh\n cd anthos-aws\n env HTTPS_PROXY=http://localhost:8118 \\\n anthos-gke aws clusters get-credentials CLUSTER_NAME\n ```\n Replace \u003cvar translate=\"no\"\u003eCLUSTER_NAME\u003c/var\u003e with your user cluster name.\n\n \u003cbr /\u003e\n\n2. Delete any [Load balancers](/kubernetes-engine/multi-cloud/docs/aws/previous-generation/how-to/loadbalancer)\n [Ingress](/kubernetes-engine/multi-cloud/docs/aws/previous-generation/how-to/ingress), or\n [PersistentVolumeClaim](https://kubernetes.io/docs/concepts/storage/persistent-volumes/#lifecycle-of-a-volume-and-claim)\n resources you created.\n\n 1. Use the following `kubectl` command to check if there are any remaining\n load balancer Services on your user cluster:\n\n env HTTPS_PROXY=http://localhost:8118 \\\n kubectl get services\n\n 2. Use the following `kubectl` command to check if there are any remaining\n PersistentVolumeClaims on your user cluster:\n\n env HTTPS_PROXY=http://localhost:8118 \\\n kubectl get persistentvolumeclaim\n\n 3. Use `kubectl delete` to delete any remaining Services or\n PersistentVolumeClaims you created. For example, to delete the\n LoadBalancer `my-lb-service`, run the following command:\n\n env HTTPS_PROXY=http://localhost:8118 \\\n kubectl delete service \u003cvar translate=\"no\"\u003emy-lb-service\u003c/var\u003e\n\n For example, to delete the\n PersistentVolumeClaims `my-pvc`, run the following command: \n\n env HTTPS_PROXY=http://localhost:8118 \\\n kubectl delete persistentvolumeclaim \u003cvar translate=\"no\"\u003emy-pvc\u003c/var\u003e\n\n3. From your `anthos-aws` directory, use\n `anthos-gke` to switch context to your management service.\n\n ```sh\n cd anthos-aws\n anthos-gke aws management get-credentials\n ```\n | **Note:** `anthos-gke` uses the credentials you specified when you ran `aws configure`.\n\n \u003cbr /\u003e\n\n4. Delete all AWSNodePools with `kubectl`.\n\n env HTTPS_PROXY=http://localhost:8118 \\\n kubectl delete AWSNodePool --all\n\n5. Delete all AWSCluster with `kubectl`.\n\n env HTTPS_PROXY=http://localhost:8118 \\\n kubectl delete AWSCluster --all\n\nRemoving a management service\n-----------------------------\n\nYou use `anthos-gke` to delete the management service.\n\n1. Change directory to the folder you created\n during [installation](/kubernetes-engine/multi-cloud/docs/aws/previous-generation/how-to/installing-management#configure).\n\n2. Use `anthos-gke` to delete resources for the management service.\n\n env HTTP_PROXY=http://localhost:8118 \\\n anthos-gke aws management delete\n\n | **Note:** The previous command uses the `HTTP_PROXY` environment variable instead of `HTTPS_PROXY`. The uninstallation process doesn't complete when using the `HTTPS_PROXY` environment variable.\n\nRemoving the AWS S3 bucket\n--------------------------\n\nWhen you create a management service, `anthos-gke` creates an S3 bucket on AWS\nwith binaries for cluster creation. The bucket is reused when you create new\nuser clusters. If you are no longer using GKE on AWS, you can\nremove the bucket after you delete your management service.\n\nTo delete the GKE on AWS S3 bucket, perform the following steps:\n\n1. To find the bucket's name, find the value of `gcs_download_bucket` in\n `terraform-workspace/terraform.tfvars.json` with `grep`.\n\n grep agent_s3_bucket terraform.tfvars.json\n\n2. Delete this s3 bucket and its contents with the `aws` command, replacing\n \u003cvar translate=\"no\"\u003ebucket-name\u003c/var\u003e with the name of your bucket from the previous\n command.\n\n aws s3 rb s3://\u003cvar translate=\"no\"\u003ebucket-name\u003c/var\u003e --force"]]