11-22-2017 08:52 AM
I'm running an extract from a SQL Server table. Before I load into my destination table, I want to remove any keys in the destination table that have a match in the work table (W8T0KH11) I produced from my extract task.
The destination table has a PK on 'TransactionKey'. If I don't run the deletion, the proc append that is run in the Table Loader errors due to the key already existing. The reason I want to delete and re-append is because the rows have been updated at source.
In the delete task, I have said where W8T0KH11.TransactionKey = DirectoryBaskets.TransactionKey. This fails when I try to run the delete stating "ERROR: Unresolved reference to table/correlation name W8T0KH11."
I'm of the opinion that for some reason the work table is not being created before I try to run the delete. From the log it is defined and then immediately moves to process delete.
I want to have these deletions run before the table loader and appreciate any pointers.
Flow screenshot :
11-23-2017 02:00 AM
11-23-2017 05:08 AM
Thanks for the response.
If you mean the Update/Insert Load style, I found this was performing individual loops per row. I have 700k+ rows to analyse and it was taking far too long (and had run up some impressive size logs).
11-23-2017 05:14 AM
.... actually looks like that was due to the technique I was using. Modify by index looks OK.