I think you're just witnessing the optimizer at work. If it thinks that
doing sequential scans is faster, it will ignore the indices.
At 11:03 AM 5/31/2001 +0200, Koen Antonissen wrote:
>Thing I descovered after i posted to the group was that after creating
>the scheme again, the indexes are used! after vacuum (analyze) the use
>of indexes was gone again on certain tables...
>
>Any other suggestions?