Try SET NOCOUNT ON at the top of the SQL Statement.

Rather than use an execute SQL task, why not write the SELECT statement in the OLEDB source component in the dataflow?

Then you use the recordcount component to populate a variable....

Hi,

If I understand you correctly then you have to put a join into your update statement something like this...

UPDATE tablename1 SET EndTime = getdate()

FROM tablename1 a

JOIN tablename2 b

ON a.columnname = b.columnname

WHERE b.columnname...

Hi Rick

I think you have to decide your granularity. There's no reason why you can't manually insert a 1 on each row in the detail fact table, allowing your users/reports...

I've come accross this problem quite a bit, and I've used 3 upside down question marks as a delimiter "¿¿¿" They've never come up in any of the wierd and...

If I understand you correctly, you are saying that because of the comma at the end of the line SSIS is automatically generating a blank column...

You should be able to...

Hi,

The work around to this is to create a blank template; use a file task to copy and overwrite your destination; then send your results to the clean file.

I've not tried this so I don't know if it works... and you may have gone down this road already.

Instead of the ODBC connection, set up an ADO.NET for ODBC connection...

I guess you would need to include the following tasks:-

1. You could use the foreach container to populate a table with filenames and properties.
2. Then sort this table in the order...

Basically, you use a lookup component to see if the record exists. If it does the record will come through the green/success output. If it doesn't it will come through...

You are describing a normalising operation which can be achieved either through the SSIS Unpivot transform or if already imported into your SQL table, you can use the UNPIVOT clause...

Its a lengthy process to describe in a post, however, to point you in the right direction, do some research on both a lookup and a merge join / conditional split.

There's...

How about using transactional replication instead. Initialise from a backup instead of a snapshot and that way you avoid the overhead on the SQL Server while the snapshot is being...

Have you tried converting to DT_DBTIMESTAMP

My apologies... I didn't read your initial post correctly.

One thing that you have to remember is that the meta data in the package can't be changed at run-time, so, if...