Have you considered just doing it in T-SQL?

INSERT INTO YourTable (C1,C2,C3)

SELECT C1,C2,C3

FROM

(Your outer join query to identify new rows)

You just run that in an execute SQL statement in SSIS...

If I understand correctly you are asking why bother with a relational star schema when you can use cubes? (or vice versa)

You've got a whole load of inconsistent dirty unaligned...

1. Managing deletes in source

Your idea is basically how replication works - you record the operation (Insert, Update, Delete) in a table in the source. Then you apply those changes...

You don't need to use SSIS, just SQL Agent.

See if this helps. Make sure you have the exact command line prepared beforehand, and make sure the SQL Agent Service account...

If you were to create a user app outside of SSIS, then you may as well remove SSIS from the equation altogether.

But this has become a design discussion and I'm...

The function MSGBox allows you to show a message and capture which button was pressed (i.e. Yes, No, Cancel). This is a VB6/VBScript function.

There would be a parallel in VB.Net...

How are you 'viewing' the saved data? It's possible its saved perfectly but your viewer doesn't support the characters you need.

You might want to be more specific - how is...

How do your environments differ? for example do your DEV, TEST, PROD environments all exist on different servers? In that case its probably easiest to use an environment variable to...

First things first. have you created a connection manager that successfully connects to your server?

This might help you along:

http://social.msdn.microsoft.com/forums/en-US/sqlintegrationservices/thread/bfbafbb9-c095-4250-9dc0-85f3731c9bda/

You need to make sure that the Sybase client is installed on...

All of the above comments boil down to: Should we use an ELT or ETL approach?

ETL being what SSIS is designed to do - transform the data in a stream...

If I could adress some points:

*Have you successfully connected to an Oracle database using that provider? The instance I need to connect to is running on a Red Hat Linux...

I am fairly certain that pk_dynreport is a package. In which case, in the following syntax:

addr.pk_dynreport.f_chk_omit('ccard_auto_renew')

addr is the schema

pk_dynreport is the package (ahem its starts with pk_)

f_chk_omit is the function...

Can you alter the the package to 'pipeline the data out'?

Then you can use this kind of syntax to get the data out:

SELECT * FROM TABLE(myschema.mypackage.myfunction());

I've read about ADA but no I've never used it. C++ was the language of choice when I was at uni (engineering not computer science though)

You know its very interesting...

If you search on a covered non-clustered index on a table with a clustered index, you get a bookmark lookup where the index goes and looks up the row in...