SURANTYN Jean François wrote:
> Many thanks for your quick reply
>
> In fact, that issue comes from a recent migration from Oracle to
> Postgresql, and even if some queries were not optimized by the past
> (example: where n=1 and n=1), Oracle was able to rewrite them and to
> "hide" the bad queries". But now that we have migrated to Postgresql,
> we have discovered that some queries were indeed badly wroten I will
> tell to the developpers to try to optimize their queries for them to
> work efficiently on Postgresql
If nothing else it will help when / if you decide to use prepared
queries - there's no way to optimise "n=$1 or n=$2" at planning time.
--
Richard Huxton
Archonet Ltd