Connecting to IBM Knowledge Catalog (IKC)
Integrating IBM® watsonx.data with IBM Knowledge Catalog (IKC) provides self-service access to data assets (tables and views) for knowledge workers who need to use the data assets to gain insights.
IBM Knowledge Catalog - watsonx.data integration supports the following connectors for governance:
- Hive
- Iceberg
- Hudi
- Delta Lake
- Oracle
- Postgresql
- Mysql
- SQL server
- Db2
Before you begin
To enable IKC integration, ensure the following pre-requisites are met:
- A working watsonx.data environment.
- A working IBM Knowledge Catalog (IKC) environment.
- IBM Knowledge Catalog and watsonx.data are configured with service-to-service authorization in IBM Cloud.
Both watsonx.data and IKC must be present in the IBM Cloud environment.
Connect and import the asset metadata from watsonx.data to IKC
-
Log in to IBM Cloud.
-
Go to Resource list > AI / Machine Learning > IBM Knowledge Catalog.
-
Click Launch in IBM Cloud Pak for Data. The IBM Knowledge Catalog home page opens.
A pop-up window opens with the options to create new catalog and category. If you prefer not to create them now, you can close the window.
-
From the left pane, go to Catalogs > View all catalogs. The Catalogs page opens with the list of available catalogs.
-
To create a new catalog:
- Click New catalog.
- On the New catalog page:
- In the Name field, add a name for the catalog.
- In the Description field, add a description.
- From the Object storage instance drop-down list, select an object storage. If you don't have one, create an instance by clicking the link in the UI.
- Select the Enforce data protection and data location rules checkbox to automatically enforce data protection and data location rules when you attempt to access data assets in the catalog.
- Toggle the Controls switch to on poistion to allow reporting on asset metadata.
- Choose an option to handle duplicate assets from the Duplicate asset handling section.
- Click Create. The catalog is created and the catalog page opens.
-
Go to Add to catalog > Connection.
-
On the New connection page, search and select IBM® watsonx.data.
-
Enter the following details:
New connection Field Description Name Enter the name of the connection. Description Enter a connection description. Connect to IBM watsonx.data on Cloud Pak for Data Do not select the checkbox. Hostname or IP address Enter the watsonx.data instance URL. For information about retrieving the Hostname, see Getting connection information. Port Enter the port number. For information about retrieving the Port, see Getting connection information. Instance ID Enter the instance ID. You can get the instance ID from the watsonx.data instance home page (information icon). Instance name Enter the watsonx.data instance name. CRN Enter the Cloud Resource Name. You can get the CRN from the watsonx.data instance home page (information icon). Username Enter your username ( ibmlhapikey_<EMAIL_ID>
).Password Enter your IAM API key. To create one, see Creating an API key. SSL is enabled Select the checkbox. SSL Certificate Download the certificate from the watsonx.data console by using a web browser and paste in this field. Engine's hostname or IP address Enter the engine hostname available in the watsonx.data console without port number and :
.Engine ID Enter the engine ID available in the watsonx.data console. Engine's port Enter the engine port number available with the engine host name. -
Optional: Click Test connection to test the connection.
-
Click Create. The connection is added to the catalog.
-
Go to Add to catalog > Connected assets.
-
In the connected asset, click Select source and navigate to the asset you want to import.
-
Select the asset and click Add. The asset is successfully added to IKC.
Configure IKC in IBM® watsonx.data UI
-
Log in to IBM® watsonx.data.
-
From the left pane, go to Access control.
-
Select the catalog to open the catalog details page.
-
Go to the Integrations tab and click Integrate service.
-
Enter the following details:
Ingrate service Field Description Service Select IBM Knowledge Catalog. Storage catalogs Select the applicable storage catalogs for IKC governance. IKC endpoint Configure the IKC API url by including api.
afterhttps://
. For example, if the URL ishttps://dataplatform.cloud.ibm.com
, the endpoint should behttps://api.dataplatform.cloud.ibm.com
. If the URL is with location (https://<region>.dataplatform.cloud.ibm.com
), the endpoint should behttps://api.<region>.dataplatform.cloud.ibm.com
. -
Click Integrate.
Verify the masking functionality as per the rules in IKC
- Login to IBM Knowledge Catalog.
- From the left pane, go to Governance > Rules.
- From the Rules page, verify that the rules corresponding to your data class of the column is defined. You can define a new rule by using Add rule button.
The owner can see the unmasked data. To verify whether masking is functioning correctly, log in to watsonx.data as user who is not the owner of the asset in IKC and query the asset.
Supported datatypes
IBM® watsonx.data IKC integration supports the following datatypes:
- Varchar
- Bigint
- Boolean
- Date
- Double
- Integer
- Smallint
- Timestamp
- Tinyint