Some components (e.g. Derived Column, Data Conversion) can create a new column from an old one and thereafter both of those columns appear in the pipeline. In many cases you are only every going to use the one that you created.
In this situation alot of people complain that they can't delete columns from the pipeline that they know that they are not going to use anymore. I used to argue that delete columns from the pipeline isn't actually possible because of the underlying buffer
architecture - but in saying that I was avoiding the problem.
Rather than delete, it would be good to "hide" columns so that IN THE DESIGNER they are not visible to downstream components (I accept that they will still be in the buffer at execution-time - that's not a problem). I have come across many situations
where being able to do this would be very very beneficial - it is very easy to get lost in the plethora of columns that can sometimes be in the pipeline - especially when some of them have identical names.
This feature really needs to added. have the develpment team tried it? It's very very annoying to have all columns come thru the pipeline without the option to turn them off.
I read this article from your forum its necessary requirement.. is there any update on this feature. its used to increase performance of ETL.
Edited byilava1990Sunday, September 22, 2013 7:37 PM
Microsoft is conducting an online survey to understand your opinion of the Technet Web site. If you choose to participate, the online survey will be presented to you when you leave the Technet Web site.