Integrate Live Databricks Data into Amazon SageMaker Canvas with RDS



Use CData Connect Cloud to connect to Databricks from Amazon RDS connector in Amazon SageMaker Canvas and build custom models using live Databricks data.

Amazon SageMaker Canvas is a no-code machine learning platform that lets you generate predictions, prepare data, and build models without writing code. When paired with CData Connect Cloud, you get instant, cloud-to-cloud access to Databricks data for building custom machine-learning models, predicting customer churn, generating texts, building chatbots, and more. This article shows how to connect to Connect Cloud from Amazon SageMaker Canvas using the RDS connector and integrate live Databricks data into your ML model deployments.

CData Connect Cloud provides a pure SQL, cloud-to-cloud interface for Databricks, allowing you to easily integrate with live Databricks data in Amazon SageMaker Canvas — without replicating the data. CData Connect Cloud looks exactly like a SQL Server database to Amazon SageMaker Canvas and uses optimized data processing out of the box to push all supported SQL operations (filters, JOINs, etc) directly to Databricks, leveraging server-side processing to quickly return Databricks data.

About Databricks Data Integration

Accessing and integrating live data from Databricks has never been easier with CData. Customers rely on CData connectivity to:

  • Access all versions of Databricks from Runtime Versions 9.1 - 13.X to both the Pro and Classic Databricks SQL versions.
  • Leave Databricks in their preferred environment thanks to compatibility with any hosting solution.
  • Secure authenticate in a variety of ways, including personal access token, Azure Service Principal, and Azure AD.
  • Upload data to Databricks using Databricks File System, Azure Blog Storage, and AWS S3 Storage.

While many customers are using CData's solutions to migrate data from different systems into their Databricks data lakehouse, several customers use our live connectivity solutions to federate connectivity between their databases and Databricks. These customers are using SQL Server Linked Servers or Polybase to get live access to Databricks from within their existing RDBMs.

Read more about common Databricks use-cases and how CData's solutions help solve data problems in our blog: What is Databricks Used For? 6 Use Cases.


Getting Started


Configure Databricks Connectivity for Amazon SageMaker Canvas

Connectivity to Databricks from Amazon SageMaker Canvas is made possible through CData Connect Cloud. To work with Databricks data from Amazon SageMaker Canvas, we start by creating and configuring a Databricks connection.

  1. Log into Connect Cloud, click Connections, and click Add Connection.
  2. Select "Databricks" from the Add Connection panel.
  3. Enter the necessary authentication properties to connect to Databricks.

    To connect to a Databricks cluster, set the properties as described below.

    Note: The needed values can be found in your Databricks instance by navigating to Clusters, and selecting the desired cluster, and selecting the JDBC/ODBC tab under Advanced Options.

    • Server: Set to the Server Hostname of your Databricks cluster.
    • HTTPPath: Set to the HTTP Path of your Databricks cluster.
    • Token: Set to your personal access token (this value can be obtained by navigating to the User Settings page of your Databricks instance and selecting the Access Tokens tab).
  4. Click Create & Test.
  5. Navigate to the Permissions tab in the Add Databricks Connection page and update the User-based permissions.

Add a Personal Access Token

If you are connecting from a service, application, platform, or framework that does not support OAuth authentication, you can create a Personal Access Token (PAT) to use for authentication. Best practices would dictate that you create a separate PAT for each service, to maintain granularity of access.

  1. Click on your username at the top right of the Connect Cloud app and click User Profile.
  2. On the User Profile page, scroll down to the Personal Access Tokens section and click Create PAT.
  3. Give your PAT a name and click Create.
  4. The personal access token is only visible at creation, so be sure to copy it and store it securely for future use.

With the connection configured, you are ready to connect to Databricks data from Amazon SageMaker Canvas.

Connecting to CData Connect Cloud from Amazon SageMaker Canvas

With the connection in CData Connect Cloud configured, you are ready to integrate live Databricks data into Amazon SageMaker Canvas using its RDS connector.

  1. Select a domain and user profile in Amazon SageMaker Canvas and click on "Open Canvas".
  2. Once the Canvas application opens, navigate to the left panel, and select "My models".
  3. Click on "Create new model" in the My models screen.
  4. Specify a Model name in Create new model window and select a Problem type. Click on "Create".
  5. Once the model version gets created, click on "Create dataset" in the Select dataset tab.
  6. In the Create a tabular dataset window, add a "Dataset name" and click on "Create".
  7. Click on the "Data Source" drop-down and search for or navigate to the RDS connector and click on " Add Connection".
  8. In the Add a new RDS connection window, set the following properties:

    • Connection Name: a relevant connection name
    • Set Engine type to sqlserver-web
    • Set Port to 14333
    • Set Address as tds.cdata.com
    • Set Username to a Connect Cloud user (e.g. [email protected])
    • Set Password to the PAT for the above user
    • Set Database name the Databricks connection (e.g., Databricks1)
  9. Click on "Create connection".

Integrating Databricks Data into Amazon SageMaker Canvas

With the connection to Connect Cloud configured in the RDS, you are ready to integrate live Databricks data into your Amazon SageMaker Canvas dataset.

  1. In the tabular dataset created in RDS with Databricks data, search for the Databricks connection configured on Connect Cloud in the search bar or from the list of connections.
  2. Select the table of your choice from Databricks, drag and drop it into the canvas on the right.
  3. You can create workflows by joining any number of tables from the Databricks connection (as shown below). Click on "Create dataset".
  4. Once the dataset is created, click on "Select dataset" to build your model.
  5. Perform analysis, generate prediction, and deploy the model.

At this point, you have access to live Databricks data in Amazon SageMaker that you can utilize to build custom ML models to generate predictive business insights and grow your organization.

SQL Access to Databricks Data from Cloud Applications

Now you have a direct connection to live Databricks data from Amazon SageMaker Canvas. You can create more connections, datasets, and predictive models to drive business — all without replicating Databricks data.

To get real-time data access to 100+ SaaS, Big Data, and NoSQL sources directly from your cloud applications, see the CData Connect Cloud.

Ready to get started?

Learn more about CData Connect Cloud or sign up for free trial access:

Free Trial