Re: Sequencing file processing

I have read on convoy processing and still have some questions.  These
files 
that I need to process are large files (can easily contain .5 mill
records). 
I do not necessarily want to process each of these messages through the 
orchestration because it might slow the overall processing down.  What I 
want to do is to use the orchestration and convoy processing to control
the 
processing of these files according to a predefined order.

The receive ****ts I have configured right now use the FILE adapter and the

maps that I have created to read the records and transform them into 
messages.  I was hoping that if the FILE adapter can publish a message out

to the message box indicating the end of the file that is being processed,

then that message could be used in a sequential convoy.  The orchestration

could enable the next receive ****t and then let that ****t process its 
respective FILE and issue an end of file message.  The orchestration can 
then disable that ****t and enable the next one and so on.  This way I can 
use the orchestration for coordination and still not have to go millions
of 
these messages through the orchestration instance.

Thanks,

Waqar



"Dan Rosanova" <dnova7@[EMAIL PROTECTED]
> wrote in message 
news:5aa59969-0cbf-48ac-b109-f5eedd750edf@[EMAIL PROTECTED]
>I think if you have the data to correlate use it.  Eric's second
> solution was probably offered for if you don't have unique correlation
> data and is a great solution for that case.  Here's an MS link for
> Convoy processing:
http://msdn2.microsoft.com/en-us/library/ms942189.aspx
> even though it says 2004, it's still 2006 applicable.
>
> Kind Regards,
> -Dan