Re: Join query on 1M row table slow - Mailing list pgsql-general

From scott.marlowe
Subject Re: Join query on 1M row table slow
Date
Msg-id Pine.LNX.4.33.0402101542090.29897-100000@css120.ihs.com
Whole thread Raw
In response to Re: Join query on 1M row table slow  (CSN <cool_screen_name90001@yahoo.com>)
Responses Re: Join query on 1M row table slow  (CSN <cool_screen_name90001@yahoo.com>)
List pgsql-general
On Tue, 10 Feb 2004, CSN wrote:

>
> I disabled enable_hashagg and enable_nestloop. Appears
> to have made both queries worse :(
>

Good, then we know that the nest loop and hash agg are probably good
plans.

>
> How exactly do I do that?
>
> SELECT * from thanks limit 1000
> ;)

it's an alter table thingie:

alter table tablename alter column columnname set statistics 100;

But since it looks like it's picking a good plan, it's probably not a real
big deal.

So, can you get rid of the join / in on the other table, or do you need
it there?


pgsql-general by date:

Previous
From: CSN
Date:
Subject: Re: Join query on 1M row table slow
Next
From: "scott.marlowe"
Date:
Subject: Re: DB cache size strategies