CODEWITHSUNDEEP
sql-serverssis
lightweight
lightweight

Reputation: 3327

task error in ssis data flow for csv import

I exported some data using sqlplus in Windows server to a csv file.

I'm trying to import the csv file into sql server using SSIS but get this error:

[OLE DB Destination [1743]] Error: An error occurred while setting up
a binding for the "TECHNOLOGY" column. The binding status was  
"DT_TEXT". The data flow column type is
"DBBINDSTATUS_UNSUPPORTEDCONVERSION". The conversion from the OLE DB
type of "DBTYPE_IUNKNOWN" to the destination column type of
"DBTYPE_VARCHAR" might not be supported by this provider.

The destination table was created using the ddl from the oracle table and in my data flow task in SSIS the varchar columns are DT_TEXT datatype. In the dest table, they are varchar(10) and in oracle, they were also varchar(10)

I tried using derived column with this expression: (DT_STR,12,1252)[Column 1]

but get this error then:

[OLE DB Destination [1770]] Error: An error occurred while setting up a binding for the     "TECHNOLOGY" column. The binding status was "DT_TEXT". The data flow column type is "DBBINDSTATUS_UNSUPPORTEDCONVERSION". The conversion from the OLE DB type of "DBTYPE_IUNKNOWN" to the destination column type of "DBTYPE_VARCHAR" might not be supported by this provider.

any ideas on this?

Upvotes: 4

Views: 8555

Answers (1)

David
David

Reputation: 1611

Right click your source transformation in the data flow and go to the advanced editor. There will be a list of columns there that shows the output of the source including the data type. Verify that it is outputting the column as text and ensure that any derived column outputs or other transformations you have along the way are not modify or replacing the data type.

If necessary, insert a derived column transformation before your destination and replace the column and force it to be text.

Also verify your mapping of column to destination column in the outbound data flow to ensure it is being mapped as you are expecting it to be.

Upvotes: 2

Related Questions