Azure ML Preview Available in Azure Germany

  • Article

This post is authored by Ted Way, Program Manager, Data Group, Microsoft.

We are excited to announce the private preview of Azure Machine Learning as part of the IOT Suite launch at Hannover Messe, the world's leading trade fair for industrial technology, in Microsoft Azure Germany. Customers interested in participating in the preview can email AzureGermany@microsoft.com. A core set of Azure ML features are available in Germany, with more in the process of being added.

What is Microsoft Azure Germany?

The German datacenter is a physically and logically separate allocation of Azure, one in which all customer data and required support systems reside in German datacenters. There is a dedicated network within Germany that is independent of the public cloud network. A German data trustee, T-Systems (Deutsche Telekom), controls physical and logical access to customer data, and Microsoft has a commitment to meet applicable compliance requirements and certifications. If you require compliance with German business and public sector requirements, then Microsoft Azure Germany would be appropriate for you. Otherwise, Azure ML is available in public Azure regions already.

What Should I Know About Microsoft Azure ML Germany?

When you provision the IOT Suite Predictive Maintenance solution through its Germany-specific page at https://www.azureiotsuite.de/, you will get an Azure ML workspace and the corresponding experiment. This is currently the only non-programmatic way to create an Azure ML workspace.

Once the workspace is created, you will be able to use Azure ML Studio as you would in public Azure. You can create new experiments, drag in modules, and run the experiment. Sample experiments are available by clicking the +New button in Studio. The model can also be operationalized as an API hosted in Microsoft Azure Germany, accessible worldwide as a REST API.

During the private preview, please note the below for the Reader and Writer modules (soon renamed Import Data and Export Data, respectively):

  • Azure SQL
    • You can connect to any Azure SQLDB instance if you fully qualify the server and instance name. In other words, do not omit the prefixes commonly associated with Azure SQLDB, but be sure use the full URL of the database.
  • Azure table
    • You can read from a table that has been made available through the SAS or Public option. Reading from any Black Forest Azure account is not supported yet.
  • Azure blob
    • You can read from a blob that has been made available through the SAS or Public option. Reading from any Black Forest Azure account is not supported yet.
  • Hive query
    • Use the HDFS option, not the Azure storage option, and reference the URI of the cluster in HDFS.
  • Data feed provider
    • Access to feeds that require authentication is not supported. Use for public data only.

How Do I Get Started?

Email AzureGermany@microsoft.com to participate in the Azure Germany preview. Documentation about Azure ML can be found here.

Ted
@tedwinway