December 28, 2005 at 5:16 pm
I receive two warning messages when I execute a particular data flow that "the output column is not subsequently used... removing this unused column can increase performance"...
"can"? seems to me that it will break the task.
In a dataflow you cannot use an OLE-DB component without a source, and the source has a query, for me it is "select stageFlag from tblTrans where stageFlag is null"... and the OLE-DB component updates the resulting rows stageFlag to 1 (true).
How can I update the stageFlag without having selected something? or am I totally missing something here?
December 30, 2005 at 4:10 am
David,
Why not just use an Execute SQL Task with the query:
UPDATE tblTrans SET stageFlag = TRUE WHERE stageFlag IS NULL
???
Does that work for you?
-Jamie
Jamie Thomson
http://sqlblog.com/blogs/jamie_thomson
January 3, 2006 at 12:03 pm
January 16, 2006 at 8:09 pm
January 16, 2006 at 11:56 pm
You've lost me. What is the difference between:
"We only want to execute this on successful completion of the previous task"
and
"not just completion was either success or failure"
???
-Jamie
Jamie Thomson
http://sqlblog.com/blogs/jamie_thomson
January 17, 2006 at 11:43 am
If the prior operation fails we do not want to update the stageFlag... only when the operation was successful do we want to do so. Doing as you suggested, the stageFlag would be updated even if we disabled the previous control flow component. There is only a success flow into the new component you suggested - and success is defined as the previous component having completed, not necessarily having completed successfully.
January 17, 2006 at 12:04 pm
Hang on, I'm lost.You say there are no OnSuccess constraints. Not only is it there - its the default.
You are talking about control-flow right?
-Jamie
Jamie Thomson
http://sqlblog.com/blogs/jamie_thomson
January 17, 2006 at 12:10 pm
January 17, 2006 at 12:37 pm
This response was intended for another message from you... must have been one of those quantum moments you hear about; but there was something else on the screen when I replied to the message from you that appears above. At any rate, doing the update your way resulted in rows being updated in the control flow even when they had failed (and thrown errors into a flat file destination) in the data flow. I returned to doing things the way I was doing them before and it works as expected. Now, I really must get off the forums and back into my own head a bit... 81 working hours until a major deliverable... thanks again Jamie. I do appreciate you very much. David
January 17, 2006 at 1:29 pm
No problem. The reason you read something that isn't there anymore is cos I misread your original post and wrote something that didn't make sense - so I got rid of it.
-Jamie
Jamie Thomson
http://sqlblog.com/blogs/jamie_thomson
Viewing 10 posts - 1 through 9 (of 9 total)
You must be logged in to reply to this topic. Login to reply