Artifact Registry 是一项支持软件包和其他格式的通用软件包管理服务。了解如何从 Container Registry 过渡,以便更灵活地控制您的工件。

版本说明

本页面记录了 Container Registry 的正式版更新。您可以定期查看本页面,以了解有关新功能或更新功能、Bug 修复、已知问题和已弃用功能的公告。

您可以在 Google Cloud 版本说明页面上查看 Google Cloud 所有产品的最新产品动态。

要接收最新产品动态,请将本页面的网址添加到您的 Feed 阅读器,或直接添加 Feed 网址:https://cloud.google.com/feeds/container-registry-release-notes.xml

February 02, 2021

On-Demand Scanning is available in Preview. You can manually scan container images stored locally on your computer or remotely in Container Registry. To get started with manual scanning, see On-Demand Scanning quickstart

October 05, 2020

Starting October 5, the Container Registry Service Agent is granted the Container Registry Service Agent IAM role by default when you enable the Container Registry API. Previously, this account was granted the Editor role.

To learn more about this change and how to update the role on an existing Container Registry service account with Editor permissions, see the Container Registry access control documentation

March 16, 2020

Artifact Registry, the evolution of Container Registry, is now available in beta. It includes support for Docker as well as Maven and npm package formats.

If you currently use Container Registry, see Transitioning from Container Registry for more information.

September 20, 2019

Container Analysis is now Generally Available.

See the documentation for information on how to scan container images for vulnerabilities as well as how to to store and retrieve metadata for container images.

August 26, 2019

V1 Version of the Container Analysis API is now Generally Available.

See the documentation for details about the API.

September 06, 2018

Container Analysis is now available in Beta.

See the documentation for information on how to store and retrieve metadata for container images in Container Registry.

November 22, 2017

Container Analysis is now available in Alpha.

See the documentation for information on how to get vulnerability scanning results.

October 26, 2017

Pub/Sub notifications feature is now available in Beta. See the documentation for more information on how to configure notifications.

March 01, 2017

V1 Version of Docker Registry API is now deprecated.

November 29, 2016

Container Registry now supports the Docker V2 image manifest, schema version 2.

August 05, 2016

The Docker client now allows users to specify a credential store, which acts as a delegate to generate the credentials that are used to authenticate requests to repositories. The Container Registry credential helper, docker-credential-gcr, simplifies authentication for all use cases, including working on a developer machine, working on a VM, and using a service account. This means you no longer have to use the Cloud SDK with Container Registry.

July 31, 2016

Added the Registry API catalog command. Container Registry now implements the catalog feature as described by the Docker Registry API specification.

June 22, 2016

Container Registry now offers a Docker Hub mirror. When a user adds the --registry-mirror=https://<us|eu|asia>-mirror.gcr.io flag to their Docker daemon, any attempt to pull an image from Docker Hub results in a call to Container Registry. If the image is found in the Container Registry mirror, it is pulled from there, and no call is made to Docker Hub. If the image is not found, the client gets a 404 result and can then make a call to Docker Hub to download the image.

June 13, 2016

The Container Registry user interface now shows Docker Registry V2 image formats and V2 image digests.

November 13, 2015

Implemented V2 of the Docker Registry API. See Docker Registry HTTP API V2.

June 22, 2015

Container Registry moved from Beta to GA.

November 12, 2014

Container Registry Beta.

Container Registry is a new service that provides private container repositories for Google Cloud Platform customers to support container development and deployment workflows. Container Registry is currently in Beta; it is suitable for experimentation and is intended to provide an early view of the production service, but customers are strongly encouraged not to run production workloads on it.