>> Can you please elaborate a bit?
> I thought that A0.Prop would ignore the composite index created on the
> columns subj and prop but this does not seem to be the case.
Yeah, I think you're barking up the wrong tree here. I think Tom had
the correct diagnosis - what do you get from "show work_mem"?
What kind of machine are you running this on? If it's a UNIX-ish
machine, what do you get from "free -m"and "uname -a"?
In response to
pgsql-performance by date
|Next:||From: Tomasz Myrta||Date: 2009-02-25 10:20:02|
|Subject: full text search - dictionary caching|
|Previous:||From: Claus Guttesen||Date: 2009-02-24 20:55:22|
|Subject: Re: Abnormal performance difference between Postgres and MySQL|