Andrew Sullivan wrote: > This is probably better on -performance, and is certainly a FAQ. > But. . . > > On Wed, Feb 23, 2005 at 03:01:52PM +0100, Thomas Braad Toft wrote: > > > > Table device contains 5285 rows, tmeevent contains 834912 rows. > ^^^^ ^^^^^^ > > > -> Seq Scan on tmeevent (cost=0.00..23606.12 rows=834912 width=138) > > (actual time=0.04..2193.97 rows=834912 loops=1) > ^^^^^^ > > > -> Seq Scan on device (cost=0.00..564.85 rows=5285 width=29) (actual > > time=0.04..25.07 rows=5285 loops=1) > ^^^^ > > > Why isn't the planner using my indexes? I tried making them as both rtree > > Because there's no advantage to using an index when you are fetching > 100% of both tables. This is the most efficient plan. Of course, > it's an open question whether you want to get 100% of both tables. > But that's what you're doing, and using the index would be more > expoensive than this.
Right. The FAQ addresses this issue. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073 ---------------------------(end of broadcast)--------------------------- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
