Issue
Migrate reports extract errors when the source has DATS and/or TIMS columns from SAP NetWeaver.
Resolution
Note: This will be resolved in a future release (DBM-3571).
In the meantime, you can work around this issue by adding manual data type conversions to the Database Types config option within Migrate.
Menu path: Administer > Advanced > Database Types
In this example, the backend DB is a SQL Server DB that is running as the SAP ECC database.
NOTE: In some cases, you may be extracting SAP Netweaver datatypes even if the DB is something like HANA or Oracle, so the mapping would need to be for SAP Netweaver to SQL Server.
Cause
Summary:
- SAP NetWeaver datatypes of DATS/TIMS map to datetime data types
- CDATA extracts DATS/TIMS as datetime, so no issue
- ODBC extracts DATS/TIMS as nvarchar, so there is a data type conversion issue during the data extraction step.
When using the data extraction components in the Migrate module of the SKP, we will usually end up with a mixed scenario for source and target systems. Most data extractions will be done with the ODBC connection if allowed. And only some tables will be extracted via the SAP NetWeaver connection.
For the System datasources, the SAP NetWeaver datatypes are coming in for both SAP Application Server and ODBC connection types.
When using snapshot management in Migrate to control which tables are extracted, the snapshot DB version of the table is created using the NetWeaver data types and the data type mapping. In particular, the NetWeaver data types DATS and TIMS are being mapped to datetime formats in SQL Server (this is likely true for Oracle & HANA working databases as well).
When using Replicate to extract from SAP NetWeaver, this does not cause any issues because the CDATA driver is converting DATS and TIMS to datetime data types. However, when extracting via ODBC, there is a data type conversion issue. ODBC extraction pulls DATS and TIMS as their native data types which is nvarchar without any data type conversion. This causes extract failures for these fields and data types.