Archives of the TeradataForum
Message Posted: Thu, 27 Sep 2001 @ 13:17:50 GMT
Well, this is a weird one I got from one of our developers. See the following >not: No, we're not doing time stamps, we're just coding the time in a CHAR(8) field >(it's kind of a "maybe we'll use this someday" field, you know source widely?), but we >are getting two different results with different multiload scripts that both just refer >to "TIME". Very strange.... [etc.]
Frank, I don't believe you've given us enough information to answer this question. I can say, though, that the handling of time fields is a little tricky. In V2R2 (?) and before, there was no separate TIME data type, so times had to be stored in some other way (in DBC, they are still stored as FLOAT). But I digress. To address your question, I remind everyone that the format of a field has nothing to do with how it's stored internally. The data type and the default format for each column are stored in DBC.TVFields; have you checked to make sure those are the same? We would need to see the relevant portions of the MultiLoad scripts too.
select drugcode, lastupdatedate, lastupdatetime from hcftest.drugmaster 00536346735 2001-09-19 1.74750 select diagnosiscode, lastupdatedate, lastupdatetime from hcftest.diagnosisicd9group 4749 2001-09-19 17:15:44
These results are indeed suspicious. Is the time in the first table defined as FLOAT?
W. Geoffrey Rommel
|Copyright 2016 - All Rights Reserved|
|Last Modified: 27 Dec 2016|