In DB2 (and other SQL databases I know) there is a clear difference between a CHAR field with empty string '' value and null (missing) value.
When using DB2 bulkload however data from the source is first put in a tab-delimited file on the DM server which uses unquoted strings, so there is no difference between the 2 values anymore.
This means an empty string in the source data is being loaded as a null value in the target, which is not the same as the source value.
Unfortunately I also have VARCHAR fields in the DB2 target table defined with NOT NULL, so I get errors on those records and they are rejected in the load, while the empty string should be accepted.
The problem seems to be that the intermediate tab-delimited file uses unqouted strings, if it uses quoted strings it would not cause this problem.
Is there some setting that DM will use quoted strings in the intermediate tab-delimited file for the bulkload?This message has been edited. Last edited by: <Kathryn Henning>,
WebFocus 8203M, iWay DataMigrator, Windows, DB2 Windows V10.5, MS SQL Server, Azure SQL, Hyperstage, ReportCaster
This looks like it's going to require some research, so I'd like to recommend opening a case on InfoResponse Online.
|Powered by Social Strata|