Exactly what I did, when I realised that there was an extra Table in the FROM with no conditions set.
Well anyway, this did clear my doubts about whether schema affects performance at all.
Robins
On 8/29/07, Robins Tharakan <robins@xxxxxxxxx> wrote:
Thanks Tom,
Exactly what I did, when I realised that there was an extra Table in the FROM with no conditions set.
Well anyway, this did clear my doubts about whether schema affects performance at all.
RobinsOn 8/28/07, Tom Lane < tgl@xxxxxxxxxxxxx> wrote:
Schemas are utterly, utterly irrelevant to performance.
I'm guessing you missed analyzing one of the tables, or forgot an index,
or something like that. Also, if you did anything "cute" like use the
same table name in more than one schema, you need to check the
possibility that some query is selecting the wrong one of the tables.
The explain output you showed is no help because the expense is
evidently down inside one of the functions in the SELECT output list.
One thing you should probably try before getting too frantic is
re-ANALYZEing all the tables and then starting a fresh session to
clear any cached plans inside the functions. If it's still slow
then it'd be worth digging deeper.
regards, tom lane
--
Robins
