Limitations for Input Data

Key Tables

Due to technical limitations, Key tables cannot be loaded in multi-threaded mode. If you decide to create input data files using the Maintenance Object method, consider the following:

• Create data extract(s) for a Maintenance Object data, but exclude the Key table(s) and run data upload in multi-threaded batches

• Create separate extract for the Key table(s) only and upload them in single threaded batches.

Duplicate Records

Duplicate records are not rejected during the upload. This may result in the index corruption and other issues in the target table(a).

The default SQL Loader data upload method configured in the system is Direct Path. Among the advantages of a Direct Path method are higher performance and more efficient resources utilization. Constraints are disabled during Direct Path load and as a result the duplicate PK records are not rejected.

There are two approaches to address the issue

• Clean up the duplicate records from the legacy data extract BEFORE the upload

• Customise the default configuration and switch to the Conventional Path load method.

  • Create custom Control File Header: Duplicate Managed Content K1-SQLLoaderControlFileHeader .and modify to use Conventional Path Load (Consult SQL Loader documentation for more detailed information).

  • Modify Conversion Instructions (Conversion Task Type): specify SQL Loader Control File Header created above

  • Regenerate Conversion Artifacts

CLOB Data Size

The maximum size of the CLOB data included in the main data file is 10K. If converted CLOB data size is expected to be larger than 10K, create secondary data files.

CLOB Data - XML

The XML data formatted using new line characters cannot be uploaded as part of the primary data file. If your CLOB data includes formatted XML, you should create secondary data file(s) for CLOB columns.