Have you tried capturing the errors? On the step which is failing, go to your error configuration manager, and choose to redirect error. Then place an OLE DB Destination, direct...

What formula did you use for your derived column? What was the data type that you chose as your output? What inputs did you use for it?

Did you try changing DT_NTEXT (UNICODE Stream), to DT_TEXT (Text Stream) ?

DBTIMESTAMP is the correct variable to pass in to a SMALLDATETIME field. See this page for more info:

http://msdn.microsoft.com/en-us/library/ms141036%28v=SQL.90%29.aspx

Check all your meta data types. Specifically, look at the meta data on...

Since your query is not actually using any of the values, you might want to use this approach instead:

SELECT h.invoice_id,

h.IssueDate

FROM invoice_head...

Well, I've found the solution. Though I'm not entirely sure on *why* this solution works.

When messing around with the project, and trying to open it on my machine and then...

Actually I should add something to this -

When I run the package from Visual Studio, it works completely fine. It's only when I try running the package by executing it...

awesome - i hate that part about SSIS. great change!

What is (DT_14) ? Are you trying to convert to an integer? If so, you should be using (DT_i4). Note that is a letter, not a number. I've put it...

for a stored proc, you'd use the same manner described in there, except it would be something like this. Let's say you had a stored proc with two input parameters...

This might help - try this:

http://www.simple-talk.com/sql/ssis/passing-variables-to-and-from-an-ssis-task/

what you probably want is all three i'm guessing. so you could do something like this:

CREATE FUNCTION dbo.uf_GetClg_Rate

(@Co As Varchar(02),

@Itemcd As Varchar(10),

@Date As Datetime)

RETURNS NUMERIC(14,4)

AS

BEGIN

...

one option which might work for you is to just import it as a raw file. in this case you won't have any delimiters, and it will just be one...

depending on the power of your machine, you could use the lookup transform instead of the merge join. lookup is *much* faster, but it also requires you to cache all...

what would probably work for you is to use two seperate connection managers to the same flat file.

in the first one, select the option to include column headers, and also...