커스텀 컨테이너를 기반으로 사용자 관리 노트북 인스턴스를 만들 수 있습니다. 커스텀 컨테이너를 사용하면 특정 니즈에 맞게 사용자 관리 노트북 환경을 맞춤설정할 수 있습니다.
컨테이너는Google Cloud 서비스 계정에서 액세스할 수 있어야 하며, 8080 포트에서 서비스를 노출해야 합니다.
Deep Learning Containers 이미지에서 파생된 컨테이너는 해당 이미지가 이미 사용자 관리 노트북과 호환되도록 구성되어 있으므로 만드는 것이 좋습니다.
커스텀 컨테이너 커널이 업데이트되는 방식
Vertex AI Workbench는 다음 상황에서 커널의 최신 컨테이너 이미지를 가져옵니다.
인스턴스를 만들 때.
인스턴스를 업그레이드할 때.
인스턴스를 시작할 때.
커스텀 컨테이너 커널은 인스턴스가 중지될 때 유지되지 않으므로 인스턴스가 시작될 때마다 Vertex AI Workbench가 컨테이너 이미지의 최신 버전을 가져옵니다.
인스턴스 실행 중에 컨테이너의 새 버전이 출시되면 인스턴스를 중지하고 시작해야 인스턴스의 커널이 업데이트됩니다.
시작하기 전에
사용자 관리형 노트북 인스턴스를 만들려면 먼저 Google Cloud 프로젝트가 있고 이 프로젝트에 Notebooks API를 사용 설정해야 합니다.
Sign in to your Google Cloud account. If you're new to
Google Cloud,
create an account to evaluate how our products perform in
real-world scenarios. New customers also get $300 in free credits to
run, test, and deploy workloads.
In the Google Cloud console, on the project selector page,
select or create a Google Cloud project.
사용자 관리형 노트북 인스턴스에 GPU를 사용하려는 경우 Google Cloud 콘솔의 할당량 페이지를 확인하여 프로젝트에 사용 가능한 GPU가 충분히 있는지 확인하세요. GPU가 할당량 페이지에 나와 있지 않거나 추가 GPU 할당량이 필요한 경우 할당량 상향 조정을 요청하세요. Compute Engine 리소스 할당량 페이지의 추가 할당량 요청을 참조하세요.
필요한 역할
프로젝트를 만든 경우 프로젝트에 대한 소유자(roles/owner) IAM 역할이 있으며 이 역할에는 모든 필수 권한이 포함됩니다. 이 섹션을 건너뛰고 사용자 관리형 노트북 인스턴스를 만듭니다. 프로젝트를 직접 만들지 않았으면 이 섹션에서 계속 진행합니다.
Vertex AI Workbench 사용자 관리형 노트북 인스턴스를 만드는 데 필요한 권한을 얻으려면 관리자에게 프로젝트에 대한 다음 IAM 역할을 부여해 달라고 요청하세요.
리전 및 영역: 새 인스턴스의 리전 및 영역을 선택합니다. 최상의 네트워크 성능을 위해 지리적으로 가장 가까운 리전을 선택합니다.
사용 가능한 사용자 관리형 노트북 위치를 참조하세요.
환경 섹션의 환경 필드에서 커스텀 컨테이너를 선택합니다.
Docker 컨테이너 이미지 필드에서 다음 방법 중 하나로 Docker 컨테이너 이미지를 추가합니다.
Docker 컨테이너 이미지 경로를 입력합니다. 예를 들어 Deep Learning Containers의 가속기로 TensorFlow 2.12 컨테이너 이미지를 사용하려면 us-docker.pkg.dev/deeplearning-platform-release/gcr.io/tf-cpu.2-12.py310을 입력합니다.
선택을 클릭하여 Artifact Registry에서 Docker 컨테이너 이미지를 추가합니다. 그런 다음 컨테이너 이미지가 저장된 Artifact Registry 탭에서 프로젝트를 컨테이너 이미지가 포함된 프로젝트로 변경한 후 컨테이너 이미지를 선택합니다.
[[["이해하기 쉬움","easyToUnderstand","thumb-up"],["문제가 해결됨","solvedMyProblem","thumb-up"],["기타","otherUp","thumb-up"]],[["이해하기 어려움","hardToUnderstand","thumb-down"],["잘못된 정보 또는 샘플 코드","incorrectInformationOrSampleCode","thumb-down"],["필요한 정보/샘플이 없음","missingTheInformationSamplesINeed","thumb-down"],["번역 문제","translationIssue","thumb-down"],["기타","otherDown","thumb-down"]],["최종 업데이트: 2025-09-04(UTC)"],[],[],null,["# Create a Vertex AI Workbench user-managed notebooks instance with a custom container\n\nCreate a user-managed notebooks instance with a custom container\n================================================================\n\n\n| Vertex AI Workbench user-managed notebooks is\n| [deprecated](/vertex-ai/docs/deprecations). On\n| April 14, 2025, support for\n| user-managed notebooks will end and the ability to create user-managed notebooks instances\n| will be removed. Existing instances will continue to function\n| but patches, updates, and upgrades won't be available. To continue using\n| Vertex AI Workbench, we recommend that you\n| [migrate\n| your user-managed notebooks instances to Vertex AI Workbench instances](/vertex-ai/docs/workbench/user-managed/migrate-to-instances).\n\n\u003cbr /\u003e\n\nYou can create a user-managed notebooks instance based on a custom\ncontainer. Using a custom container lets you customize a\nuser-managed notebooks environment for your specific needs.\nThe container must be accessible to your\nGoogle Cloud service account and expose a service on port 8080.\nWe recommend creating a container derived from a\n[Deep Learning Containers\nimage](/deep-learning-containers/docs/choosing-container#choose_a_container_image_type),\nbecause those images are already configured to be compatible\nwith user-managed notebooks.\n\nHow custom container kernels are updated\n----------------------------------------\n\nVertex AI Workbench pulls the latest container image for your kernel:\n\n- When you create your instance.\n\n- When you upgrade your instance.\n\n- When you start your instance.\n\nThe custom container kernel doesn't persist when your instance is stopped,\nso each time your instance is started, Vertex AI Workbench pulls\nthe latest version of the container image.\n\nIf your instance is running when a new version of a container is released,\nyour instance's kernel isn't updated until you stop and start your instance.\n\nBefore you begin\n----------------\n\nBefore you can create a user-managed notebooks instance, you must have a Google Cloud project and enable the Notebooks API for that project.\n\n- Sign in to your Google Cloud account. If you're new to Google Cloud, [create an account](https://console.cloud.google.com/freetrial) to evaluate how our products perform in real-world scenarios. New customers also get $300 in free credits to run, test, and deploy workloads.\n- In the Google Cloud console, on the project selector page,\n select or create a Google Cloud project.\n\n | **Note**: If you don't plan to keep the resources that you create in this procedure, create a project instead of selecting an existing project. After you finish these steps, you can delete the project, removing all resources associated with the project.\n\n [Go to project selector](https://console.cloud.google.com/projectselector2/home/dashboard)\n-\n [Verify that billing is enabled for your Google Cloud project](/billing/docs/how-to/verify-billing-enabled#confirm_billing_is_enabled_on_a_project).\n\n-\n\n\n Enable the Notebooks API.\n\n\n [Enable the API](https://console.cloud.google.com/flows/enableapi?apiid=notebooks.googleapis.com&redirect=https://console.cloud.google.com)\n\n- In the Google Cloud console, on the project selector page,\n select or create a Google Cloud project.\n\n | **Note**: If you don't plan to keep the resources that you create in this procedure, create a project instead of selecting an existing project. After you finish these steps, you can delete the project, removing all resources associated with the project.\n\n [Go to project selector](https://console.cloud.google.com/projectselector2/home/dashboard)\n-\n [Verify that billing is enabled for your Google Cloud project](/billing/docs/how-to/verify-billing-enabled#confirm_billing_is_enabled_on_a_project).\n\n-\n\n\n Enable the Notebooks API.\n\n\n [Enable the API](https://console.cloud.google.com/flows/enableapi?apiid=notebooks.googleapis.com&redirect=https://console.cloud.google.com)\n\n1. If you plan to use GPUs with your user-managed notebooks instance, [check the quotas page in the\n Google Cloud console](https://console.cloud.google.com/quotas) to ensure that you have enough GPUs available in your project. If GPUs are not listed on the quotas page, or you require additional GPU quota, you can request a quota increase. See [Requesting an increase in\n quota](/compute/quotas#requesting_additional_quota) on the Compute Engine [Resource quotas](/compute/quotas) page.\n\n\u003cbr /\u003e\n\n### Required roles\n\nIf you created the project, you have the\nOwner (`roles/owner`) IAM role on the project,\nwhich includes all required permissions. Skip this section and\nstart creating your user-managed notebooks instance. If you didn't\ncreate the project yourself, continue in this section.\n\n\nTo get the permissions that\nyou need to create a Vertex AI Workbench user-managed notebooks instance,\n\nask your administrator to grant you the\nfollowing IAM roles on the project:\n\n- Notebooks Admin ([`roles/notebooks.admin`](/vertex-ai/docs/workbench/user-managed/iam#notebooks.admin))\n- Service Account User ([`roles/iam.serviceAccountUser`](/iam/docs/understanding-roles#iam.serviceAccountUser))\n\n\nFor more information about granting roles, see [Manage access to projects, folders, and organizations](/iam/docs/granting-changing-revoking-access).\n\n\nYou might also be able to get\nthe required permissions through [custom\nroles](/iam/docs/creating-custom-roles) or other [predefined\nroles](/iam/docs/roles-overview#predefined).\n\n### Make sure your custom container is ready\n\nMake sure you have a custom container that is accessible to your\nGoogle Cloud service account. For information about how to create a\ncustom container from a\n[Deep Learning Containers image](/deep-learning-containers/docs/choosing-container#choose_a_container_image_type), see\n[Creating a derivative container](/deep-learning-containers/docs/derivative-container).\n\nCreate an instance with a custom container\n------------------------------------------\n\nTo create a user-managed notebooks instance\nwith a custom container, complete the following steps:\n\n1. In the Google Cloud console, go to the **User-managed notebooks** page.\n Or go to [notebook.new](https://notebook.new)\n (https://notebook.new) and skip the next step.\n\n [Go to User-managed notebooks](https://console.cloud.google.com/vertex-ai/workbench/user-managed)\n2. Click add_box **Create new**.\n\n3. Click **Advanced options**.\n\n4. On the **Create instance** page,\n in the **Details** section,\n provide the following information for your new instance:\n\n - **Name**: a name for your new instance\n - **Region** and **Zone** : Select a region and zone for the new instance. For best network performance, select the region that is geographically closest to you. See the available [user-managed notebooks\n locations](/vertex-ai/docs/general/locations#user-managed-notebooks-locations).\n5. In the **Environment** section, in the **Environment** field,\n select **Custom container**.\n\n6. In the **Docker container image** field, add a Docker container image\n in one of the following ways:\n\n - Enter a Docker container image path. For example, to use a TensorFlow 2.12 container image with accelerators from [Deep Learning Containers](/deep-learning-containers/docs/choosing-container#deciding), enter `us-docker.pkg.dev/deeplearning-platform-release/gcr.io/tf-cpu.2-12.py310`.\n - Click **Select** to add a Docker container image from Artifact Registry. Then on the **Artifact Registry** tab where your container image is stored, change the project to the project that includes your container image, and select your container image.\n7. Make the rest of your selections, or leave them on their default\n setting. For more information about these settings, see [Create a\n user-managed notebooks instance with specific properties](/vertex-ai/docs/workbench/user-managed/create-new#create-with-options).\n\n8. Click **Create**. Vertex AI Workbench creates\n a user-managed notebooks instance for you, based\n on your custom container.\n\nWhat's next\n-----------\n\n- Read about how to [push container images to\n Artifact Registry](/artifact-registry/docs/docker/pushing-and-pulling). If the container images you push to Artifact Registry are derived from a [Deep Learning Containers\n image](/deep-learning-containers/docs/choosing-container#choose_a_container_image_type), you can use these container images when creating user-managed notebooks instances.\n- Learn more about modifying your custom containers by reading [Best practices for writing\n Dockerfiles](https://docs.docker.com/develop/develop-images/dockerfile_best-practices/)."]]
