Re: Really really slow select count(*) - Mailing list pgsql-performance

From Craig Ringer
Subject Re: Really really slow select count(*)
Date
Msg-id 4D50868C.1020209@postnewspapers.com.au
Whole thread Raw
In response to Re: Really really slow select count(*)  (Greg Smith <greg@2ndquadrant.com>)
List pgsql-performance
On 02/08/2011 03:05 AM, Greg Smith wrote:

> Accordingly I would expect any serious
> attempt to add some auto-reduction behavior to be beset with argument,
> and I'd never consider writing such a thing as a result. Too many
> non-controversial things I could work on instead.

Yep. I expressed my own doubts in the post I suggested that in.

If Pg did auto-correct down, it'd be necessary to scream about it
angrily and continuously, not just once during startup. Given that it's
clear many people never even look at the logs ("what logs? where are
they?") I think Pg would also have to send notices to the client.
Problem is, many clients don't process notices/warnings, so particularly
slack admins won't see that either.

I'm not particularly excited about the idea.

--
Craig Ringer

pgsql-performance by date:

Previous
From: Craig Ringer
Date:
Subject: Re: Really really slow select count(*)
Next
From: Sylvain Rabot
Date:
Subject: Indexes with condition using immutable functions applied to column not used