In this article, the author discusses improvements in the B-tree index based on sorting duplicate key values using CTID.