Hi,

I have a question:

I have a DTS package with two tasks:
1. A Copy SQL Server Object Task: to copy a table object between two
database servers
2. A Execute SQL Task: to delete all rows in the table previously copied

This is a database used by a web application, and there might be inserts
into the table between these two steps.
How can I garantee that the rows that are deleted in step 2 are only those
copied in step 1, and none that are newly inserted?
I've have read about the transaction isolation level that I can define in
the DTS package, namely about
"Repeatable Read. You are guaranteed not to see any changes made by other
transactions in values it has already read."
Does this work in my case?

Thanks for your help
Rafaela

The problem here being that the transfer is one transaction and the delete
another.

You could Pump the records over (You do not need to recreate the table
everytime surely)
You could then have a linked server back to the source that says delete from
source where the record PK = ones I just copied over.



--

----------------------------
Allan Mitchell (Microsoft SQL Server MVP)
MCSE,MCDBA
www.SQLDTS.com
I support PASS - the definitive, global community
for SQL Server professionals - http://www.sqlpass.org



"RKAzinhal" <rafaelaka (AT) hotmail (DOT) com> wrote

The problem is, the table has no PK, because it's a log table...
Are there other suggestions?

Thanks, Rafaela

"Allan Mitchell" <allan (AT) no-spam (DOT) sqldts.com> wrote

Give it a PK then. an integer with identity will do. Without a PK or unique
constraint you will struggle to identify a record.

--

----------------------------
Allan Mitchell (Microsoft SQL Server MVP)
MCSE,MCDBA
www.SQLDTS.com
I support PASS - the definitive, global community
for SQL Server professionals - http://www.sqlpass.org



"RKAzinhal" <rafaelaka (AT) hotmail (DOT) com> wrote