How to connect and process IBM Cloud Object Storage data from Azure Databricks

Databricks is a cloud-based service that provides data processing capabilities through Apache Spark. When paired with the CData JDBC Driver, customers can use Databricks to perform data engineering and data science on live IBM Cloud Object Storage data. This article explains how to host the CData JDBC Driver in Azure, as well as connect to and process live IBM Cloud Object Storage data in Databricks.

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.

Install the CData JDBC Driver in Azure

To work with live IBM Cloud Object Storage data in Databricks, install the driver through Azure Data Lake Storage (ADLS). (Please note that the method of connecting through DBFS, which previous versions of this article described, has been deprecated, but has not published an end-of-life.)

1. Upload the JDBC JAR file to a blob container of your choice (i.e. "jdbcjars" container of the "databrickslibraries" storage account).
2. Fetch the Account Key from the storage account by expanding "Security + networking" and clicking on "Access Keys". Show and copy whichever of the two keys you wish to use. 👁 Get Access Key
   
3. Get the JDBC JAR file's URL by navigating to Containers, opening the specific container storing the JAR, and selecting the entry for the JDBC JAR file. This should open the file's details, where there should be a convenient button to copy the URL button to clipboard. This value will look similar to the below, though the "blob" component may vary depending on storage account type:

   https://databrickslibraries.blob.core.windows.net/jdbcjars/cdata.jdbc.salesforce.jar

   👁 Get JAR URL
   
4. In the Configuration tab of your Databricks cluster, click on the Edit button and expand "Advanced options". From there, add the following Spark option (derived from the JAR URL's domain name) with your copied Account key as its value and click Confirm: spark.hadoop.fs.azure.account.key.databrickslibraries.blob.core.windows.net 👁 Apply Account Key
   
5. In the Libraries tab of your Databricks cluster, click on "Install new", and select the ADLS option. Specify the ABFSS URL for the driver JAR (also derived from the JAR URL's domain name), and click Install. The ABFSS URL should resemble the below:

   abfss://[email protected]/cdata.jdbc.salesforce.jar

   👁 Install ADLS Library
   

Connect to IBM Cloud Object Storage from Databricks

With the JAR file installed, we are ready to work with live IBM Cloud Object Storage data in Databricks. Start by creating a new notebook in your workspace. Name the workbook, make sure Python is selected as the language (which should be by default), click on Connect and under General Compute select the cluster where you installed the JDBC driver (should be selected by default).

👁 Attaching to an existing compute resource

Configure the Connection to IBM Cloud Object Storage

Connect to IBM Cloud Object Storage by referencing the class for the JDBC Driver and constructing a connection string to use in the JDBC URL. Additionally, you will need to set the property in the JDBC URL (unless you are using a Beta driver). You can view the licensing file included in the installation for information on how to set this property.

driver = "cdata.jdbc.ibmcloudobjectstorage.IBMCloudObjectStorageDriver"
url = "jdbc:ibmcloudobjectstorage:RTK=5246...;ApiKey=myApiKey;CloudObjectStorageCRN=MyInstanceCRN;Region=myRegion;OAuthClientId=MyOAuthClientId;OAuthClientSecret=myOAuthClientSecret;"

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.

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

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, 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.

👁 Using the built-in connection string designer to generate a JDBC URL (Salesforce is shown.)

Load IBM Cloud Object Storage Data

Once the connection is configured, you can load IBM Cloud Object Storage data as a dataframe using the CData JDBC Driver and the connection information.

remote_table = spark.read.format ( "jdbc" ) \
	.option ( "driver" , driver) \
	.option ( "url" , url) \
	.option ( "dbtable" , "Objects") \
	.load ()

Display IBM Cloud Object Storage Data

Check the loaded IBM Cloud Object Storage data by calling the display function.

display (remote_table.select ("Key"))

👁 Displaying IBM Cloud Object Storage Data

Analyze IBM Cloud Object Storage Data in Azure Databricks

If you want to process data with Databricks SparkSQL, register the loaded data as a Temp View.

remote_table.createOrReplaceTempView ( "SAMPLE_VIEW" )

The SparkSQL below retrieves the IBM Cloud Object Storage data for analysis.

result = spark.sql("SELECT Key, Etag FROM SAMPLE_VIEW WHERE Bucket = 'someBucket'")

The data from IBM Cloud Object Storage is only available in the target notebook. If you want to use it with other users, save it as a table.

remote_table.write.format ( "parquet" ) .saveAsTable ( "SAMPLE_TABLE" )

👁 Displaying IBM Cloud Object Storage Data

Download a free, 30-day trial of the CData JDBC Driver for IBM Cloud Object Storage and start working with your live IBM Cloud Object Storage data in Azure Databricks. Reach out to our Support Team if you have any questions.