Validate Hybrid Partitioned Data
To validate a hybrid partitioned table, you can use the
                procedure DBMS_CLOUD.VALIDATE_HYBRID_PART_TABLE. This
                procedure includes a parameter that lets you specify a specific partition to
                validate.
               
Before validating a hybrid partitioned table you need to create the table. To create
                                a hybrid partitioned table use the procedure DBMS_CLOUD.CREATE_HYBRID_PART_TABLE
                                (see Query Hybrid Partitioned Data for more details):
                  
BEGIN
 DBMS_CLOUD.VALIDATE_HYBRID_PART_TABLE(
    table_name => 'HPT1',
    partition_name => 'P1');
END;
/
This procedure scans your source files for partition P1 and validates them using the format options specified when you create the hybrid partitioned table.
The validation of a hybrid partitioned table by default
                                validates all the external partitions sequentially until
                                        rowcount is reached. If you specify a
                                        partition_name then only that specific
                                partition is validated.
                  
The validate operation, by default, scans all the rows in your source files and stops
                                when a row is rejected. If you want to validate only a subset of the
                                rows, use the rowcount parameter. When the
                                        rowcount parameter is set the validate
                                operation scans rows and stops either when a row is rejected or when
                                the specified rowcount number of rows are validated
                                without errors.
                  
For example, the following validate operation scans 100 rows and stops when a row is rejected or when 100 rows are validated without errors:
BEGIN 
  DBMS_CLOUD.VALIDATE_HYBRID_PART_TABLE (
     table_name => 'HPT1',
     rowcount=>100 ); 
END; 
/If you do not want the validate to stop when a row is rejected and you want to see
                                all rejected rows, set the stop_on_error parameter
                                to FALSE. In this case DBMS_CLOUD.VALIDATE_HYBRID_PART_TABLE
                                scans all rows and reports all rejected rows.
                  
If you want to validate only a subset of rows use the rowcount parameter. When rowcount is set and stop_on_error is set to FALSE, the validate operation scans rows and stops either when the specified number of rows are rejected or when the specified number of rows are validated without errors. For example, the following example scans 100 rows and stops when 100 rows are rejected or when 100 rows are validated without errors:
                  
BEGIN 
  DBMS_CLOUD.VALIDATE_HYBRID_PART_TABLE (
     table_name => 'HPT1',
     rowcount => 100 
     stop_on_error => FALSE );
END; 
/For detailed information about DBMS_CLOUD.VALIDATE_HYBRID_PART_TABLE
                                parameters see VALIDATE_HYBRID_PART_TABLE Procedure.
                  
See View Logs for Data Validation to see the results of validate operations in
                                the tables dba_load_operations and
                                        user_load_operations.
                  
Parent topic: Query External Data with Autonomous AI Database