The July 2026 release of Data Ingestion and Replication includes the following changed behavior.
Task configuration wizard enhancement
After you select a destination connection or source connection for an application or database ingestion and replication task in the task configuration wizard, the connection name appears after the page name in the row of navigation buttons at the top for easy reference, for example, Destination: <connection_name> or Source: <connection_name>.
Note:
If you add a new connection in the wizard, the connection name doesn't appear at the top until you click the newly added box for it on the Destination or Source page.
Nulls for unmodified Oracle columns can be skipped during change data replication to Databricks
Database ingestion and replication incremental load jobs that have an Oracle source and Databricks target and that have Oracle supplemental logging enabled on primary key columns only can now be configured to skip replicating null values for the before images and after images of any unmodified source columns to the target. This new behavior is available for Standard and Soft Deletes apply modes only. To enable this behavior, set the target custom property writeSkipUpdatingUnmodifiedColumns to true. By default, the property value is false to preserve pre-existing behavior. However, with the pre-existing behavior, null values for the unmodified columns might not be properly handled during Update processing, leading to the source and target becoming out of sync.
Note:
If you enable skipping nulls for unmodified columns and a primary key column in a source table changes during CDC processing, the table is placed into an error state to prevent the job from failing on restart. To restart the job and recover the table, use the Resume with Options > Resync command.