Azure Machine Learning Studio - Import from Cosmos Db

Question

I am currently trying to connect to a DocumentDb (MongoDb) using Azure Machine Learning Studio.

I am currently following this guide, however it seems out of date already. The assumptions I have taken have lead me to get an Error 1000: ... DocumentDb client threw an exception The underlying connection was closed. The connection was closed unexpectedly.

The guide, and Azure Machine Learning Studio, outline the following parameters to make a connection.

Endpoint URL, Database ID, DocumentDb Key, Collection ID. It also tells you to look under the Keys blade to find these, which does not exist anymore.

These are the assumptions I have taken;

• Endpoint URL = host + port under the Connection String blade. https://host.com:port/
• Database ID = the database name listed under the Data Explorer blade.
• DocumentDb Key = Primary Password under the Connection String blade.
• Collection ID = the name of a collection in the db from the Data Explorer blade.

I have, for now, also opened all connections to the database just to make sure I wasn't closing the network to outside requests which, I guess, means that at least the DocumentDb key is a poor assumption.

---

After some input from Jon, below, here is the current state of things

• Endpoint URL = the Uri from the Overview blade.
• Database ID = the database name listed under the Data Explorer blade.
• DocumentDb Key = the Primary Password under the Connection String blade.
• Collection ID = the name of a collection in the db from the Data Explorer blade.
• Sql query = select top 10 * from CollectionID
• Sql parameters = {}

Answer 1

Through discussion in the comments, it may be that the "Endpoint URL" just needed to be updated, but I'll go over all of the inputs in case anyone else needs a reference to it.

• Endpoint URL - Can use the URI in the CosmosDB "Overview" pane in the Azure Portal
• Database ID - The name of the database to connect to
• DocumentDB Key - The primary password from the "Connection Strings" pane in the Azure Portal
• Collection ID - The name of the collection to read data from

And, for reference, here's what my data explorer looks like in CosmosDB (database ID then collection ID):

And the settings in Azure ML Studio to import the data: