Re: BUG #15954: Unable to alter partitioned table to set logged

Поиск
Список
Период
Сортировка
От David G. Johnston
Тема Re: BUG #15954: Unable to alter partitioned table to set logged
Дата
Msg-id CAKFQuwZHFOQDcx1tnyjV=kT9ss5dPDk8AKbCPv3ughhUJh5yHg@mail.gmail.com
обсуждение исходный текст
Ответ на Re: BUG #15954: Unable to alter partitioned table to set logged  (Kyotaro Horiguchi <horikyota.ntt@gmail.com>)
Ответы Re: BUG #15954: Unable to alter partitioned table to set logged  (Michael Paquier <michael@paquier.xyz>)
Список pgsql-bugs
On Mon, Apr 22, 2024 at 12:29 PM Kyotaro Horiguchi <horikyota.ntt@gmail.com> wrote:
At Wed, 21 Aug 2019 22:44:34 +0900, Michael Paquier <michael@paquier.xyz> wrote in <20190821134434.GA9383@paquier.xyz>
> On Wed, Aug 21, 2019 at 09:24:38AM -0400, Keith Fiske wrote:
> > As the others have stated, I think we just need to get some sort of
> > consistent method of handling this. Currently running an ALTER TABLE on the
> > parent to change the UNLOGGED state simply does nothing which is not
> > intuitive whatsoever. Even if it's just throwing an error saying you cannot
> > change this property, that would be better until a more thorough solution
> > can be implemented in the future.
>
> One problem with an error is that it may break existing application
> code :(

It is quite strange that we can CREATE both LOGGED and UNLOGGED
partitioned table but cannot ALTER the property. I believe no one
does ALTER TABLE SET (UN)LOGGED expecting it is silently ignored.


I have no issue leaving the unlogged aspect of an unlogged partitioned table being basically something the system ignores.  But it is a bug to fail to update the metadata related to this if it is explicitly altered.  Can we at least get that part fixed?

There is little upside to preventing the existence of an unlogged partitioned table at this point.  I have just submitted a doc patch to deal with that side of the equation.

David J.

В списке pgsql-bugs по дате отправления:

Предыдущее
От: Richard Guo
Дата:
Сообщение: Re: BUG #18442: Unnecessary Sort operator in indexScan Plan
Следующее
От: David Rowley
Дата:
Сообщение: Re: BUG #18442: Unnecessary Sort operator in indexScan Plan