Unless your package is executing a sql statement using linked servers against either A or B with no data flow present then yes it is flowing from A => C...

file seems to be missing a valid value for customerkey - when you preview the data on the source component does it show you values for the customerkey column?
If...

Partitioning will most likely not help here.

where exactly is the contention happening?
And are server specs up to scratch? sometimes performance is due to server either being too...

create one foreach loop for the folders based on a record set created by a script
then within the foreach do another foreach for each file within that folder.

If you can't do explain plans better ask your DBA's to give you that - it is a must for any developer and there is no reason not to be...

In order to allow us to help you we will need the explain plan of the execution of the query.

tables DDL and indexes will also help.

In...

As Phil mentioned IDENTITY and SEQUENCE are the most common and easier to work with .

If not feasible something like this will be the next best thing e.g....

For monitoring I do recomend SQL Monitor - using it for years and can't go without it.

It has a fully functional 14 days evaluation which may be...

If you get the explain plans for both executions (even just the estimated will suffice) you will most likely see the reason for it.

If I get this right...

Check the the linked server properties and ensure they are the same on both servers.
And I am assuming you are using the oledb provider, not the ODBC one, on...

What are the times for each one of the steps
1 - teradata to staging
2 - staging to final db
3 - final db to tabular (load data)
4...

Most likely the merge is doing data movement.

I recall a document that highlighted that issue very well but could not find it - but as a general rule...

Hum.

Have you looked at the query timeout property of the linked server itself? if this one is set it will be used instead of the global setting.
select...

From my own experience on doing these type of changes my advise is to create another table with the desired datatypes and then do batch inserts from the source table...