da-zero (4/16/2010)

---

Ha, of course you're right.

I divised my suggestion from the template that I always use. This template has a conditional split to sort out the insert and the updates....

da-zero (4/16/2010)

---

I think you can just use an OLE DB Source to read the data....

OK Rookie, I'll go easy on you 🙂

If a matching record is found (based on ID), do you just want to ignore it? Or perform some sort of update?

We need a bit more info ...

What exactly do you mean by import? Do you have matching PKs? What happens if a match is found - update or ignore? Should...

da-zero (4/15/2010)

---

Two hours today that is 🙂

(I don't count the hours constructing the package,...

da-zero (4/15/2010)

---

<rant>

I just wasted 2 hours on Excel in SSIS, so I can definitely tell you what the disadvantages are 🙂

---

And don't get me started on the automatic conversion of...

Max worksheet size is 1,048,576 rows by 16,384 columns in Excel 2007.

The fact that there is no explicit column data-typing in Excel is the single biggest cause of problems in...

Easiest thing to do is post some sample data. Your description is not detailed enough.

Creation, development, production then deletion.

Unhealthy SSIS packages skip the production stage.

AAAV (4/14/2010)

---

ok by mistake i created the dtsx in sql server 2008 BIDS and trying to open in 2005.

my dev machine has both versions and the server has only 2005....

No idea about the NULLs datatype, sorry.

Regarding SQL Server 2008 - go for it. SSIS is not much different, but in some areas it is just plain better and more...

da-zero (4/14/2010)

---

That's why I talked about redirecting or ignoring the error output...

I am wondering whether the fact that you used file-level copy with the packages has confused SSIS because of the duplicate IDs all over the place.

You may like to try...

Sorg should be joined to source_id, not destination_id.

The rest should be ok ... as long as you do not need sorg later on in the dataflow.

Build a mapping table (OldID, NewID) and then use a Lookup transformation to lookup the new ID based on the old ID ... much easier than building a mega-complicated nested...