Adam Ruth wrote:
> The optimizer is looking at the statistics and figuring that the second
> query could be done better with a sequential scan (perhaps there aren't
> many rows). Have you analyzed the table to get the statistics up to
> date? It could also be that the seq scan is faster with the size of
> your table.
Thanks! An analyze fixed it and reduced our search time from seconds
to milliseconds.
Is there ever any reason to do a "vacuum" without doing a "vacuum analyze"?
Mark