That sure seems to bolster the theory that performance is degrading because you exhaust the cache space and need to start reading index pages. When inserting sequential data, you don't need to randomly access pages all over the index tree. -Kevin >>> Kelly Burkhart <kelly@xxxxxxxxxxxxxxxxxxx> >>> I modified my original program to insert generated, sequential data. The following graph shows the results to be flat: <http://kkcsm.net/pgcpy_20051111_1.jpg> ---------------------------(end of broadcast)--------------------------- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
