แก้ไข

แชร์ผ่าน


Azure Machine Learning feature availability across clouds regions

Learn what Azure Machine Learning features are available in the Azure Government, Azure Germany, and Microsoft Azure operated by 21Vianet regions.

In the list of global Azure regions, there are several regions that serve specific markets in addition to the public cloud regions. For example, the Azure Government and the Azure operated by 21Vianet regions. Azure Machine Learning is deployed into the following regions, in addition to public cloud regions:

  • Azure Government regions US-Arizona and US-Virginia.
  • Azure operated by 21Vianet region China-East-2.

Azure Machine Learning is still in development in air-gap Regions.

For the Italy North Region, Application Insights is not available until 12/12/2023 without allow list. This will effect the following service until then - Job schedule - Feature store - Model monitor - Data import

The information in the rest of this document provides information on what features of Azure Machine Learning are available in these regions, along with region-specific information on using these features.

Azure Government

Feature Public cloud status US-Virginia US-Arizona
Automated machine learning
Create and run experiments in notebooks GA YES YES
Create and run experiments in studio web experience Public Preview YES YES
Industry-leading forecasting capabilities GA YES YES
Support for deep learning and other advanced learners GA YES YES
Large data support (up to 100 GB) Public Preview YES YES
Azure Databricks integration GA NO NO
SQL, Azure Cosmos DB, and HDInsight integrations GA YES YES
Machine Learning pipelines
Create, run, and publish pipelines using the Azure Machine Learning SDK GA YES YES
Create pipeline endpoints using the Azure Machine Learning SDK GA YES YES
Create, edit, and delete scheduled runs of pipelines using the Azure Machine Learning SDK GA YES* YES*
View pipeline run details in studio GA YES YES
Create, run, visualize, and publish pipelines in Azure Machine Learning designer GA YES YES
Azure Databricks Integration with ML Pipeline GA NO NO
Create pipeline endpoints in Azure Machine Learning designer GA YES YES
Integrated notebooks
Workspace notebook and file sharing GA YES YES
R and Python support GA YES YES
Virtual Network support GA YES YES
Compute instance
Managed compute Instances for integrated Notebooks GA YES YES
Jupyter, JupyterLab Integration GA YES YES
Virtual Network (VNet) support GA YES YES
Configure Apache Spark pools to perform data wrangling Public Preview No No
SDK support
Python SDK support GA YES YES
Security
Managed virtual network support Preview Preview Preview
Virtual Network (VNet) support for training GA YES YES
Virtual Network (VNet) support for inference GA YES YES
Scoring endpoint authentication Public Preview YES YES
Workplace private endpoint GA GA GA
ACI behind VNet Public Preview NO NO
ACR behind VNet GA YES YES
Private IP of AKS cluster Public Preview NO NO
Network isolation for managed online endpoints GA NO NO
Compute
quota management across workspaces GA YES YES
Kubernetes compute GA NO NO
Data for machine learning
Create, view, or edit datasets and datastores from the SDK GA YES YES
Create, view, or edit datasets and datastores from the UI GA YES YES
View, edit, or delete dataset drift monitors from the SDK Public Preview YES YES
View, edit, or delete dataset drift monitors from the UI Public Preview YES YES
Machine learning lifecycle
Model profiling (SDK/CLI v1) GA YES PARTIAL
The Azure Machine Learning CLI v1 GA YES YES
Visual Studio Code integration Public Preview NO NO
Event Grid integration Public Preview NO NO
Integrate Azure Stream Analytics with Azure Machine Learning Public Preview NO NO
Labeling images and text
Labeling Project Management Portal GA YES YES
Labeler Portal GA YES YES
Labeling using private workforce GA YES YES
ML assisted labeling (Image classification and object detection) Public Preview YES YES
Responsible ML
Explainability in UI Public Preview NO NO
Differential privacy SmartNoise toolkit OSS NO NO
Custom tags in Azure Machine Learning to implement datasheets GA NO NO
Fairness Azure Machine Learning Integration Public Preview NO NO
Interpretability SDK GA YES YES
Training
Experimentation log streaming GA YES YES
Experimentation UI Public Preview YES YES
.NET integration ML.NET 1.0 GA YES YES
Inference
Managed online endpoints GA YES YES
Batch inferencing GA YES YES
Other
Open Datasets Public Preview YES YES
Custom Azure AI Search (SDK v1) Public Preview YES YES

Azure Government scenarios

Scenario US-Virginia US-Arizona Limitations
General security setup
Disable/control internet access (inbound and outbound) and specific VNet PARTIAL PARTIAL
Placement for all associated resources/services YES YES
Encryption at-rest and in-transit. YES YES
Root and SSH access to compute resources. YES YES
Maintain the security of deployed systems (instances, endpoints, etc.), including endpoint protection, patching, and logging PARTIAL PARTIAL ACI behind VNet currently not available
Control (disable/limit/restrict) the use of ACI/AKS integration PARTIAL PARTIAL ACI behind VNet currently not available
Azure role-based access control (Azure RBAC) - Custom Role Creations YES YES
Control access to ACR images used by ML Service (Azure provided/maintained versus custom) PARTIAL PARTIAL
General Machine Learning Service Usage
Ability to have a development environment to build a model, train that model, host it as an endpoint, and consume it via a webapp YES YES
Ability to pull data from ADLS (Data Lake Storage) YES YES
Ability to pull data from Azure Blob Storage YES YES

Other Azure Government limitations

  • For Azure Machine Learning compute instances, the ability to refresh a token lasting more than 24 hours is not available in Azure Government.

  • Model Profiling does not support 4 CPUs in the US-Arizona region.

  • Sample notebooks may not work in Azure Government if it needs access to public data.

  • IP addresses: The CLI command used in the required public internet access instructions does not return IP ranges. Use the Azure IP ranges and service tags for Azure Government instead.

  • For scheduled pipelines, we also provide a blob-based trigger mechanism. This mechanism is not supported for CMK workspaces. For enabling a blob-based trigger for CMK workspaces, you have to do extra setup. For more information, see Trigger a run of a machine learning pipeline from a Logic App (SDK/CLI v1).

  • Firewalls: When using an Azure Government region, add the following hosts to your firewall setting:

    • For Arizona use: usgovarizona.api.ml.azure.us
    • For Virginia use: usgovvirginia.api.ml.azure.us
    • For both: graph.windows.net

Azure operated by 21Vianet

Feature Public cloud status CH-East-2 CH-North-3
Automated machine learning
Create and run experiments in notebooks GA YES N/A
Create and run experiments in studio web experience Preview YES N/A
Industry-leading forecasting capabilities GA YES N/A
Support for deep learning and other advanced learners GA YES N/A
Large data support (up to 100 GB) Preview YES N/A
Azure Databricks Integration GA YES N/A
SQL, Azure Cosmos DB, and HDInsight integrations GA YES N/A
Machine Learning pipelines
Create, run, and publish pipelines using the Azure Machine Learning SDK GA YES N/A
Create pipeline endpoints using the Azure Machine Learning SDK GA YES N/A
Create, edit, and delete scheduled runs of pipelines using the Azure Machine Learning SDK GA YES N/A
View pipeline run details in studio GA YES N/A
Create, run, visualize, and publish pipelines in Azure Machine Learning designer GA YES N/A
Azure Databricks Integration with ML Pipeline GA YES N/A
Create pipeline endpoints in Azure Machine Learning designer GA YES N/A
Integrated notebooks
Workspace notebook and file sharing GA YES N/A
R and Python support GA YES N/A
Virtual Network support GA YES N/A
Compute instance
Managed compute Instances for integrated Notebooks GA YES N/A
Jupyter, JupyterLab Integration GA YES N/A
Managed virtual network support Preview Preview N/A
Virtual Network (VNet) support GA YES N/A
SDK support
Python SDK support GA YES N/A
Security
Virtual Network (VNet) support for training GA YES N/A
Virtual Network (VNet) support for inference GA YES N/A
Scoring endpoint authentication Preview YES N/A
Workplace Private Endpoint GA YES N/A
ACI behind VNet Preview NO N/A
ACR behind VNet GA YES N/A
Private IP of AKS cluster Preview NO N/A
Network isolation for managed online endpoints GA NO N/A
Compute
quota management across workspaces GA YES N/A
Kubernetes compute GA NO NO
Data for machine learning
Create, view, or edit datasets and datastores from the SDK GA YES N/A
Create, view, or edit datasets and datastores from the UI GA YES N/A
View, edit, or delete dataset drift monitors from the SDK Preview YES N/A
View, edit, or delete dataset drift monitors from the UI Preview YES N/A
Machine learning lifecycle
Model profiling GA YES N/A
The Azure DevOps extension for Machine Learning & the Azure Machine Learning CLI GA YES N/A
FPGA-based Hardware Accelerated Models Deprecating Deprecating N/A
Visual Studio Code integration Preview NO N/A
Event Grid integration Preview YES N/A
Integrate Azure Stream Analytics with Azure Machine Learning Preview NO N/A
Labeling
Labeling Project Management Portal GA YES N/A
Labeler Portal GA YES N/A
Labeling using private workforce GA YES N/A
ML assisted labeling (Image classification and object detection) Preview YES N/A
Responsible AI
Explainability in UI Preview NO N/A
Differential privacy SmartNoise toolkit OSS NO N/A
custom tags in Azure Machine Learning to implement datasheets GA YES N/A
Fairness Azure Machine Learning Integration Preview NO N/A
Interpretability SDK GA YES N/A
Training
Experimentation log streaming GA YES N/A
Reinforcement Learning Deprecating Deprecating N/A
Experimentation UI GA YES N/A
.NET integration ML.NET 1.0 GA YES N/A
Inference
Managed online endpoints GA YES N/A
Batch inferencing GA YES N/A
Azure Stack Edge with FPGA Deprecating Deprecating N/A
Other
Open Datasets Preview YES N/A
Custom Azure AI Search Preview YES N/A

Other Azure operated by 21Vianet limitations

  • Azure operated by 21Vianet has limited VM SKU, especially for GPU SKU. It only has NCv3 family (V100).

  • REST API Endpoints are different from global Azure. Use the following table to find the REST API endpoint for Azure operated by 21Vianet regions:

    REST endpoint Global Azure China-Government
    Management plane https://management.azure.com/ https://management.chinacloudapi.cn/
    Data plane https://{location}.experiments.azureml.net https://{location}.experiments.ml.azure.cn
    Microsoft Entra ID https://login.microsoftonline.com https://login.chinacloudapi.cn
  • Sample notebook may not work, if it needs access to public data.

  • IP address ranges: The CLI command used in the required public internet access instructions does not return IP ranges. Use the Azure IP ranges and service tags for Microsoft Azure operated by 21Vianet instead.

  • Azure Machine Learning compute instances preview is not supported in a workspace where Private Endpoint is enabled for now, but CI will be supported in the next deployment for the service expansion to all Azure Machine Learning regions.

  • Searching for assets in the web UI with Chinese characters will not work correctly.

Next steps

To learn more about the regions that Azure Machine Learning is available in, see Products by region.