Re: Trigger performance

From: pginfo <pginfo(at)t1(dot)unisoftbg(dot)com>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org>
Subject: Re: Trigger performance
Date: 2004-01-22 15:31:34
Message-ID: 400FECD6.B835AF9C@t1.unisoftbg.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Hi,

thanks for the answer.
It is very interest, because I readet many times that if I write the trigger
in "C" it will work faster.
In wich case will this trigger work faster if write it in "C"?
In all my triggres I have "select ...." or "insert into mytable select ..."
or "update mytable set ...where...".
I need this info because I have a table with ~1.5 M rows and if I start to
update 300 K from this rows it takes ~ 2h.
If I remove the trigger for this table all the time is ~ 1 min.

regards,
ivan.

Tom Lane wrote:

> pginfo <pginfo(at)t1(dot)unisoftbg(dot)com> writes:
> > I was supprised that the pgsql trigger take ~8 sec. to insert this rows
> > and the "C" trigger take ~ 17 sec.
>
> The reason is that plpgsql caches the plan for the invoked SELECT,
> whereas the way you coded the C function, it's re-planning that SELECT
> on every call.
>
> regards, tom lane
>
> ---------------------------(end of broadcast)---------------------------
> TIP 2: you can get off all lists at once with the unregister command
> (send "unregister YourEmailAddressHere" to majordomo(at)postgresql(dot)org)

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Tom Lane 2004-01-22 15:37:54 Re: Trigger performance
Previous Message postgres 2004-01-22 15:13:23 Function & Update/Insert Problem and Performance