Greg Stark wrote:
All that said clustering is indeed often quite effective. Especially if it makes an index scan efficient enough to win over sequential scans you can see some huge effects. It's most useful for tables that aren't undergoing lots of updates and don't need to be reclustered often.
Which is my situation; most tables are read-only. Thanks for the comments.
-Kevin ---------------------------(end of broadcast)--------------------------- TIP 2: Don't 'kill -9' the postmaster
