I was reading the Stairway on Indexes. For the tables that I am loading in chunks by year, would I be better off creating a clustered index that has the date in the first part of the index to facilitate the load and delete of the table? Or should I set the clustered index to use columns that report queries may more often use and create a nonclustered index on the date? I understand that the second option would cause the database to reorder rows during the load.