the batch/bunch approach for deletions rows it is very interesting. I want to avoid to manage a cursor for a such operation.
In my scenario, I need to populate a working table that represents the data source for a SSAS solution: I want to avoid to partition a working table.
In order to SELECT ... INTO, DROP old table and rename the new table, ALTER permissions are necessary.
My working table could countain 100 millions of records: 70 millions of rows could be deleted and maintained 30 millions of rows. In this case, using a temp table to fill is a good solution? However I could need to delete 40 millions of records and maintain 60 millions of records.
Ultimately, the batch approach seems the better solution ..., isn't it?