This topic describes techniques applicable to a Classic Machine Learning Web service.
When you create Web services that you sell forward to your customers, you need to provide trained models to each customer that are still linked to the experiment from which the Web service was created. In addition, any updates to the experiment should be applied selectively to an endpoint without overwriting the customizations.
To accomplish this, Azure Machine Learning allows you to create multiple endpoints for a deployed Web service. Each endpoint in the Web service is independently addressed, throttled, and managed. Each endpoint is a unique URL and authorization key that you can distribute to your customers.
Try Azure Machine Learning for free
No credit card or Azure subscription required. Get started now.
Adding endpoints to a Web service
There are two ways to add an endpoint to a Web service.
- Through the Azure Machine Learning Web Services portal
Once the endpoint is created, you can consume it through synchronous APIs, batch APIs, and excel worksheets. In addition to adding endpoints through this UI, you can also use the Endpoint Management APIs to programmatically add endpoints.
If you have added additional endpoints to the Web service, you cannot delete the default endpoint.
Adding an endpoint programmatically
You can add an endpoint to your Web service programmatically using the AddEndpoint sample code.
Adding an endpoint using the Azure Machine Learning Web Services portal
- In Machine Learning Studio, on the left navigation column, click Web Services.
- At the bottom of the Web service dashboard, click Manage endpoints. The Azure Machine Learning Web Services portal opens to the endpoints page for the Web service.
- Click New.
- Type a name and description for the new endpoint. Endpoint names must be 24 character or less in length, and must be made up of lower-case alphabets or numbers. Select the logging level and whether sample data is enabled. For more information on logging, see Enable logging for Machine Learning Web services.