メインコンテンツへスキップ

​
Prerequisites

Before you can configure and use Databricks with Connect AI, you must first connect a data source to your Connect AI account. See Sources for more information. You must also generate a Personal Access Token (PAT) on the Settings page. Copy this down, as it acts as your password during authentication.

​
Connect to Connect AI

To establish a connection from Databricks to Connect AI, follow these steps.
1
Download and install the Connect AI JDBC driver.
1
Open the Integrations page of Connect AI.
2
Search for JDBC or Databricks.
3
Click Download and select your operating system.
4
When the download is complete, run the setup file.
5
When the installation is complete, the JAR file can be found in the installation directory.
2
Log in to Databricks.
3
In the navigation pane, select Compute. Start any compute or create a new one.
4
Once the compute is started, click the compute and then select the Libraries tab.
5
Click Install new. The Install library dialog appears.
6
Select DBFS. Then drag and drop the JDBC JAR file into the indicated area. The file has the name cdata.jdbc.connect.jar. Click Install.
7
You must now run three notebook scripts, one by one.
8
The first script is below. Change the following:
  • Update User with your Connect AI username.
  • Update Password with the PAT you generated in the prerequisites.
  • Update Your_Connection_name with the name of the data source you created in the prerequisites.
driver = "cdata.jdbc.connect.ConnectDriver"
url ="jdbc:connect:AuthScheme=Basic;User=user@cdata.com;Password=***********;URL=https://cloud.cdata.com/api/;DefaultCatalog= Your_Connection_Name;"
9
Run the first script.
10
From the menu on the right side, select Add cell below to add a second script. The second script is below. Change the following:
  • Update User with your Connect AI username.
  • Update Password with the PAT you generated in the prerequisites.
  • Update Your_Connection_name with the name of the data source you created in the prerequisites.
  • Update YOUR_SCHEMA.YOUR_TABLE with your schema and table, for example, PUBLIC.CUSTOMERS.
remote_table = spark.read.format ( "jdbc" ) \
.option ( "driver" , "cdata.jdbc.connect.ConnectDriver") \
.option ( "url","jdbc:connect:AuthScheme=Basic;User=user@cdata.com;Password=*******;URL=https://cloud.cdata.com/api/;DefaultCatalog= Your_Connection_Name;") \
.option ( "dbtable" , "YOUR_SCHEMA.YOUR_TABLE") \
.load ()
11
Run the second script.
12
Add a cell for the third script. The third script is below. Select the columns you want to display.
display (remote_table.select ("ColumnName1","ColumnName2"))
13
Run the third script.
14
You can preview your data in Databricks.