Brian Wipf <[EMAIL PROTECTED]> writes:
> I'm trying to determine why an identical query is running
> approximately 500 to 1000 times slower on our production database
> compared to our backup database server.
It looks to me like it's pure luck that the query is fast on the backup
server. The
I added to the exists query qualifier: AND cp.category_id = 1000962
(in addition to the cp.category_id = ac.category_id)
Now I am getting a much better query plan on our production server:
Index Scan using attribute_category__category_id_fk_idx on
attribute_category ac (cost=0.00..485.71 ro
I'm trying to determine why an identical query is running
approximately 500 to 1000 times slower on our production database
compared to our backup database server.
Both database servers are dual 2.3 GHz G5 Xserves running PostgreSQL
8.1.3; both are configured with 8GB of RAM with identical