I´m importing a large XML-file with SSIS and have performanceproblem.
I have 5 output collections from my generated XSD but I´m just using
one of them, I have tried to uncheck the output columns in the unused
output collections in XML Source Editor but then I get a error messege
when I execute the package. I have also tried to remove the outputs
from Advanced Editor but get a errormessage telling me that a output
can´t be removed.

Any suggestions how to get rid of the unwanted collections?

Regards, Kuno

Hello Kuno,

What was the error you got and also how do you know ot what makes you think
that this is the issue?


Allan

The error I get when I uncheck the output columns in the unused output
collections is a Validation error "output contains no output columns.
An asynchcronous output must contain output columns"
The reason why I think this is the issue is that when I execute the
package I get warnings that the columns, that I´m now trying to
remove, "is not subsequently used in the Data Flow task. Removing this
unused output column can increase Data Flow task performance."

/Kuno


Allan Mitchell wrote:

Hello Kuno,


Ok So you have outputs created by the component, this is fine. Async has
to contain columns (they use a different buffer and so you must explicitly
give it columns).

In the UI You cannot remove Outputs
You can also not remove the columns

Whilst the warning is true about removing unused output columns from the
pipeline I am not sure how detrimental this will be. The reason i say that
is you are not going to be carrying these rows further does the line. They
stop right where they are because you do not use that output.



Allan

Thanks Allan for the response. The problem seems to be even though I'm
not using 4 of the connections they steal a lot of memory, the package
runs and completes but it is very slow.

Any suggestion how to make it faster.

Kuno

Allan Mitchell wrote:

Hello Kuno,


I would make sure that the reason for the slowness is what you think it is.
Is the pipeline complex? How do you know it is this part of the pipeline
that is slow.

We have written a component which may help you see where the slowness in
the pipeline exists.


Row Count Plus Transformation
(http://www.sqlis.com/default.aspx?77)


Allan