Thx for yours helps.
But Isn't there a possibility that a backend has more than one
transaction with locks?
Cedric.
Tom Lane wrote:
> Oliver Jowett <oliver@opencloud.com> writes:
>
>>(later) A bit of experimentation yields this (I'm sure there's a better
>>way):
>> select transaction from pg_locks where pid = pg_backend_pid();
>
>
> You'd need to qualify that more, since as-is it will also show rows for
> other locks the transaction holds. Possibly "where transaction is not
> null" will be sufficient.
>
> regards, tom lane
>
> ---------------------------(end of broadcast)---------------------------
> TIP 6: Have you searched our list archives?
>
> http://archives.postgresql.org
>
--
Cédric Coulon cedric.coulon@lina.univ-nantes.fr
Atlas group, INRIA
LINA - Université de Nantes voice: +33 (0)2 51 12 59 63
2, Rue de la Houssinière BP 92208 fax : +33 (0)2 51 12 58 97
44322 Nantes Cedex 03 France