Dear All

we recently put in a DTS request to move 36 million records from a non SQL
Server to ours. At 33 million it failed, which was a bit annoying as it had
been running about 19 hours.

Is there a way of stopping it from rollbacking the records ?

Thanks
J

On the options tab, you can set a batch size commit point. If you set
it to 10,000 then every 10,000 records will be committed. If the
package fails, you will only lose the records that have been imported
since the last batch commit point. But this usually slows things down a
little, and I wouldn't use a commit point of ten thousand - bump it up a
bit.

But why would you not want to rollback if it fails?
Why would you want an incomplete set?

Simon Worth

Julie wrote: