Build AI/ML Models with Live IBM Cloud Object Storage Data using Dataiku

Dataiku is a data science and machine learning platform used for data preparation, analysis, visualization, and AI/ML model deployment, enabling collaborative and efficient data-driven decision-making. When paired with the CData JDBC Driver for IBM Cloud Object Storage, Dataiku enhances data integration, preparation, real-time analysis, and reliable model deployment for IBM Cloud Object Storage data.

With built-in optimized data processing, the CData JDBC Driver offers unmatched performance for interacting with live IBM Cloud Object Storage data. When you issue complex SQL queries to IBM Cloud Object Storage, the driver pushes supported SQL operations, like filters and aggregations, directly to IBM Cloud Object Storage and utilizes the embedded SQL engine to process unsupported operations client-side (often SQL functions and JOIN operations). Its built-in dynamic metadata querying allows you to work with and analyze IBM Cloud Object Storage data using native data types.

This article shows how you can easily integrate to IBM Cloud Object Storage using CData JDBC Driver for IBM Cloud Object Storage in Dataiku DSS (Data Science Studio) platform, allowing you to prepare the data and build custom AI/ML models.

Preparing the Dataiku DSS environment

In this section, we will explore how to set up Dataiku, as previously introduced, with IBM Cloud Object Storage data. Be sure to install Dataiku DSS (On-Prem version) for your preferred operating system, beforehand.

Install the CData JDBC Driver for IBM Cloud Object Storage

First, install the CData JDBC Driver for IBM Cloud Object Storage on the same machine as Dataiku. The JDBC Driver will be installed in the following path:

Connecting the JDBC Driver in Dataiku DSS

To use the CData JDBC driver in Dataiku, you must create a new SQL database connection and add the JDBC Driver JAR file in the DSS connection settings.

1. Log into Dataiku DSS platform. It should open locally in your browser (e.g. localhost:11200)
2. Click on Navigate to other sections of Dataiku menu on the top right section of the platform and select Administration.
3. Select the Connections tab.
4. In Connections, click on New Connections button.
5. Now, scroll down and select Other SQL databases.

6. Generate a JDBC URL for connecting to IBM Cloud Object Storage, beginning with jdbc:ibmcloudobjectstorage: followed by a series of semicolon-separated connection string properties.

   Register a New Instance of Cloud Object Storage

   If you do not already have Cloud Object Storage in your IBM Cloud account, follow the procedure below to install an instance of SQL Query in your account:

   1. Log in to your IBM Cloud account.
   2. Navigate to the page, choose a name for your instance and click Create. You will be redirected to the instance of Cloud Object Storage you just created.

   Connecting using OAuth Authentication

   There are certain connection properties you need to set before you can connect. You can obtain these as follows:

   API Key

   To connect with IBM Cloud Object Storage, you need an API Key. You can obtain this as follows:

   1. Log in to your IBM Cloud account.
   2. Navigate to the Platform API Keys page.
   3. On the middle-right corner click "Create an IBM Cloud API Key" to create a new API Key.
   4. In the pop-up window, specify the API Key name and click "Create". Note the API Key as you can never access it again from the dashboard.

   Cloud Object Storage CRN

   If you have multiple accounts, you will need to specify the CloudObjectStorageCRN explicitly. To find the appropriate value, you can:

   • Query the Services view. This will list your IBM Cloud Object Storage instances along with the CRN for each.
   • Locate the CRN directly in IBM Cloud. To do so, navigate to your IBM Cloud Dashboard. In the Resource List, Under Storage, select your Cloud Object Storage resource to get its CRN.

   Connecting to Data

   You can now set the following to connect to data:

   • InitiateOAuth: Set this to GETANDREFRESH. You can use InitiateOAuth to avoid repeating the OAuth exchange and manually setting the OAuthAccessToken.
   • ApiKey: Set this to your API key which was noted during setup.
   • CloudObjectStorageCRN (Optional): Set this to the cloud object storage CRN you want to work with. While the connector attempts to retrieve this automatically, specifying this explicitly is recommended if you have more than Cloud Object Storage account.

   When you connect, the connector completes the OAuth process.

   1. Extracts the access token and authenticates requests.
   2. Saves OAuth values in OAuthSettingsLocation to be persisted across connections.

   Built-in Connection String Designer

   For assistance in constructing the JDBC URL, use the connection string designer built into the IBM Cloud Object Storage JDBC Driver. Either double-click the JAR file or execute the jar file from the command-line.

   java -jar cdata.jdbc.ibmcloudobjectstorage.jar

   Fill in the connection properties and copy the connection string to the clipboard.

   A typical JDBC URL is given below:

   jdbc:ibmcloudobjectstorage:ApiKey=myApiKey;CloudObjectStorageCRN=MyInstanceCRN;Region=myRegion;OAuthClientId=MyOAuthClientId;OAuthClientSecret=myOAuthClientSecret;
7. On the New SQL database (JDBC) connection screen, enter a name in the New connection name field and specify the basic parameters:
   • JDBC Driver Class: cdata.jdbc.ibmcloudobjectstorage.IBMCloudObjectStorageDriver
   • JDBC URL: JDBC connection URL obtained in the previous step
   • Driver jars directory: the folder path where the JAR file is installed on your system

   Next, select the SQL dialect of your choice. Here, we have selected 'SQL Server' as the preferred dialect. Click on Create. If the connection is successful, a prompt will display, saying 'Connection OK'.

8. The Data Catalog window will appear. Select the desired connection, catalog, and schema from the Connection to browse, Restrict to catalog, and Restrict to schema dropdowns, then click on List Tables. The Dataiku platform will list all the required tables.
9. Select any table from the list and click Preview to view the table data. Click Close to exit the window.

Creating a new project

To prepare data flows, create dashboards, analyze the IBM Cloud Object Storage data, and build AI and ML models in the Dataiku DSS platform, you need to first create a new project.

1. Select Projects from the Navigate to other sections of Dataiku menu.
2. In the Projects screen, click New Project and select + Blank Project.
3. In the New Project window, assign a Name and Project Key. Click Create. The new project dashboard opens up.
4. Select Notebooks from the menu at the top of the project screen.
5. Click on + Create Your First Notebook dropdown menu and select Write your own option.
6. In the New Notebook window, select SQL.
7. Now, select the required connection from the Connection dropdown and enter a name in the Notebook Name field.

Testing the connection

To test the IBM Cloud Object Storage connection and analyze the IBM Cloud Object Storage data, write a query in the query compiler and click Run. The queried/filtered IBM Cloud Object Storage data results will then appear on the screen.

Get Started Today

Download a free, 30-day trial of the CData JDBC Driver for IBM Cloud Object Storage to integrate with Dataiku, and effortlessly build custom AI/ML models from IBM Cloud Object Storage data.

Reach out to our Support Team if you have any questions.