pentaho/pentaho-kettle · error · KettleException

No fields defined to load to database

Error message

No fields defined to load to database

What it means

The PostgreSQL bulk loader step builds a COPY command from the mapped stream fields, and none were defined. getCopyCommand checks meta.getFieldStream() and aborts the transformation when the field mapping is empty, because a COPY without a column list is invalid for this step's generated SQL.

Solutions

  1. Open the PGBulkLoader step dialog, go to the Fields tab, and click 'Get Fields' to populate the stream-to-table field mapping
  2. Verify the transformation file/repository actually saved the field mappings (check the ktr XML for field entries under the step)
  3. Ensure upstream steps produce fields before configuring the loader, then re-save the transformation
  4. Repair the source transformation/step metadata so PGBulkLoaderMeta.getFieldStream() returns a non-empty array

Example fix

// before (empty mapping in step meta)
String[] streamFields = meta.getFieldStream(); // null or length 0 -> KettleException
// after (configure fields programmatically or via dialog)
meta.setFieldStream( new String[] { "id", "name" } );
meta.setFieldTable( new String[] { "id", "name" } );
Defensive patterns

Strategy: validation

Validate before calling

String[] streamFields = meta.getFieldStream();
if ( streamFields == null || streamFields.length == 0 ) {
  throw new IllegalStateException( "PGBulkLoader has no field mapping; open the Fields tab and use Get Fields" );
}

Try / catch

try {
  step.processRow();
} catch ( KettleException e ) {
  if ( e.getMessage().contains( "No fields defined" ) ) {
    // reconfigure field mapping before retry
  }
}

Prevention

When it happens

Trigger: Calling getCopyCommand (via copyCmd) when the step's field mapping table has no rows: meta.getFieldStream() returns null or an empty array, typically because the transformation was never fully configured or the field mappings were cleared.

Common situations: Importing a transformation XML or repository export where the field mapping did not persist; a 'Get Fields' was never clicked in the step dialog; upstream steps renamed/removed fields so the mapping grid saved empty; hand-editing the ktr and deleting the <fields> entries.

Understand the failure class

Background: "must not be empty", "cannot be empty" — required-field validation errors across open-source libraries — this error's family across 41 libraries.

Related errors

AI-assisted analysis of pentaho/pentaho-kettle@f3058517a1 (2026-09-13). Data as JSON: /api/errors/eceab1c27c1a01a3. Report an issue: GitHub.

Appendix: source

Thrown at plugins/postgresql-db-bulk-loader/impl/src/main/java/org/pentaho/di/trans/steps/pgbulkloader/PGBulkLoader.java:109

    //
    // contents.append("SET DATESTYLE ISO;"); // This is the default but we set it anyway...
    // contents.append(Const.CR);

    // Create a Postgres / Greenplum COPY string for use with a psql client
    contents.append( "COPY " );
    // Table name

    contents.append( tableName );

    // Names of columns

    contents.append( " ( " );

    String[] streamFields = meta.getFieldStream();
    String[] tableFields = meta.getFieldTable();

    if ( streamFields == null || streamFields.length == 0 ) {
      throw new KettleException( "No fields defined to load to database" );
    }

    for ( int i = 0; i < streamFields.length; i++ ) {
      if ( i != 0 ) {
        contents.append( ", " );
      }
      contents.append( dm.quoteField( tableFields[i] ) );
    }

    contents.append( " ) " );

    // The "FROM" filename
    contents.append( " FROM STDIN" ); // FIFO file

    // The "FORMAT" clause
    contents.append( " WITH CSV DELIMITER AS '" ).append( environmentSubstitute( meta.getDelimiter() ) )
        .append( "' QUOTE AS '" ).append(
      environmentSubstitute( meta.getEnclosure() ) ).append( "'" );

View on GitHub (pinned to f3058517a1)