 |
 |
Archives of the TeradataForum
Message Posted: Tue, 24 Aug 2010 @ 09:13:49 GMT
| Subj: | | Re: SysMaxRow size? |
| |
| From: | | Victor Sokovin |
| | Also, good point about DBC dqbl -vs- PMCP dbql data type definitions. The SQL test in DBC is UNICODE (as you say) but the one in the
history PMCP table is LATIN!!! However, I can't think of any UNICODE data being saved in that column - its' all SQL we write for ETL that is
'snapped' by the system periodically. Also, as we are a US company only and don't have UNICODE data, it's unlikely that any could get in
there. | |
In this case, UNION will cause conversion of the Latin column to Unicode, and that's where you likely hit the limit because byte count doubles
with this type of conversion.
As you seem confident about the data try to explicitly convert the UNICODE column to Latin in your UNION query. It would be really strange if
you hit the row size limit in that scenario.
Victor
| |
