From: | Ivan Voras <ivoras(at)freebsd(dot)org> |
---|---|
To: | pgsql-performance(at)postgresql(dot)org |
Subject: | Re: View vs Stored Proc Performance |
Date: | 2009-09-15 15:14:06 |
Message-ID: | h8oave$ohn$1@ger.gmane.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
Merlin Moncure wrote:
> On Sat, Sep 12, 2009 at 7:51 AM, Dimitri Fontaine
> <dfontaine(at)hi-media(dot)com> wrote:
>> Merlin Moncure <mmoncure(at)gmail(dot)com> writes:
>>> like joining the result to another table...the planner can see
>>> 'through' the view, etc. in a function, the result is fetched first
>>> and materialized without looking at the rest of the query.
>> I though the planner would "see through" SQL language functions and
>> inline them when possible, so they often can make for parametrized
>> views...
>
> It can happen for simple functions but often it will not. For views
> it always happens.
Are functions in language 'sql' handled differently than those of
language 'plpgsql'?
I think they're not so in any case a function will behave as a black box
with regards to the planner and optimizer (and views are always
'transparent').
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2009-09-15 15:26:13 | Re: View vs Stored Proc Performance |
Previous Message | Kevin Grittner | 2009-09-15 14:50:10 | Re: How to post Performance Questions |