December 13, 2013 at 9:23 am
I've run into an issue where I'm getting data conversion errors on a test Flat File. The Flat File connection contains the same column names and datatypes as the file being imported however the columns were in a different order for this one particular test file.
It seems to me that the FlatFile Connection data validation isn't really based on Column names but position of the column itself. Does anyone know if that's actually the issue?
For example:
FlatFile Connection
CompanyName DT_STR
ID DT_I4
Amount DT_Numeric
Flat File to be imported
ID DT_I4
CompanyName DT_STR
Amount DT_Numeric
Conversion error on CompanyName DT_STR to DT_I4 instead of it automatically mapping to same named column.
December 13, 2013 at 9:27 am
Once the definition is set I don't believe you can change it at run-time. So if two columns are reversed for example it will probably break. And if that happens I have to question WHY it is allowed. Are you allowing willy nilly file creation? Why the variation?
CEWII
December 13, 2013 at 9:33 am
Elliott,
Thanks for the reply. I just ran across it in a test scenario. I'm using a script task to identify the file based on the content and then use conditional control flow to push it to the correct data flow task. These two files are almost identical b/c they come from the same company. There is just a slight variation between the two (mainly just a different order of the columns) and when I kept getting a conversion error on this particular column I didn't understand why b/c the data was exactly the same. I suspected column order was the issue.
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply