Thread: Partitioning triggers doc patch

Attached is a doc patch based on a suggestion by Rural Hunter to
address something he found confusing while setting up partitioning.

Original thread is here:

http://archives.postgresql.org/pgsql-admin/2012-03/msg00066.php

-Kevin

On Mon, Mar 12, 2012 at 10:41 AM, Kevin Grittner
<Kevin.Grittner@wicourts.gov> wrote:
> Attached is a doc patch based on a suggestion by Rural Hunter to
> address something he found confusing while setting up partitioning.
>
> Original thread is here:
>
> http://archives.postgresql.org/pgsql-admin/2012-03/msg00066.php
>

you can create trigger that execute before the trigger that redirect
to the partition, for example to stamp the time or the user... so it
should state something like i said about constraints

"""
Do not define any check constraints on this table, unless you intend
them to be applied equally to all partitions.
"""

--
Jaime Casanova         www.2ndQuadrant.com
Professional PostgreSQL: Soporte 24x7 y capacitación

Jaime Casanova <jaime@2ndquadrant.com> wrote:
> On Mon, Mar 12, 2012 at 10:41 AM, Kevin Grittner
> <Kevin.Grittner@wicourts.gov> wrote:
>> Attached is a doc patch based on a suggestion by Rural Hunter to
>> address something he found confusing while setting up
>> partitioning.
>>
>> Original thread is here:
>>
>> http://archives.postgresql.org/pgsql-admin/2012-03/msg00066.php
>>  
> 
> you can create trigger that execute before the trigger that
> redirect to the partition, for example to stamp the time or the
> user... so it should state something like i said about constraints
> 
> """
> Do not define any check constraints on this table, unless you
> intend them to be applied equally to all partitions.
> """
Yeah, I thought about that issue, but the problem is that the rule
or trigger that routes the operation to another partition might
prevent the trigger from firing at all.  I figured that if I went as
far as what you describe, I would need to get into those timing
issues, too; and I wasn't sure whether that might be "too much
information" for that particular spot.  But if you think we should
go there, I could draft something which pointed out that an AFTER
trigger is never useful and that a BEFORE trigger to enforce things
globally must sort ahead of the "routing" trigger, and should not be
used at all if a RULE is used to route operations to the partitions.
Do we really want to get into all that in the middle of step-by-step
instructions for the simple case?
-Kevin

On Mon, Mar 12, 2012 at 12:39 PM, Kevin Grittner
<Kevin.Grittner@wicourts.gov> wrote:
>
> Do we really want to get into all that in the middle of step-by-step
> instructions for the simple case?
>

is there any other place we can add a more detailed explanation? so
from your text after saying "usually there is no reason for" you can
add a parentesis saying "more info in <link>"

--
Jaime Casanova         www.2ndQuadrant.com
Professional PostgreSQL: Soporte 24x7 y capacitación

Jaime Casanova <jaime@2ndquadrant.com> wrote:
> is there any other place we can add a more detailed explanation?
> so from your text after saying "usually there is no reason for"
> you can add a parentesis saying "more info in <link>"
Sure, I'll look around for a good spot.
-Kevin