From: | Robert Haas <robertmhaas(at)gmail(dot)com> |
---|---|
To: | Gianni Ciolli <gianni(dot)ciolli(at)2ndquadrant(dot)it> |
Cc: | postgres hackers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Dump EXTENSION sequences too |
Date: | 2012-04-24 13:23:34 |
Message-ID: | CA+TgmoaH=sMHH22rsGwVnx+shOJJiArSyVzUeyCOQPz4n5XydQ@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Thu, Apr 19, 2012 at 6:42 AM, Gianni Ciolli
<gianni(dot)ciolli(at)2ndquadrant(dot)it> wrote:
> currently an EXTENSION can mark some of its tables as "configuration
> tables" using pg_catalog.pg_extension_config_dump(), so that pg_dump
> "does the right thing".
>
> I think it would be useful to mark sequences too, but unfortunately it
> is not possible; hence, each time a dump is reloaded, all the
> sequences in the extension are reset to 1, causing all the related
> problems.
>
> Moreover, the error message that we get if we try to mark a sequence
> does not mention the requirement that the relation is a table. The
> "OID %u does not refer to a table" error message seems to be wrongly
> restricted to the case when get_rel_name can't find a relation.
>
> Is there any objection to the above proposal? I did a little search of
> the archives, but I couldn't find any related discussions; I apologise
> if I missed something.
Seems like a good idea to me.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
From | Date | Subject | |
---|---|---|---|
Next Message | Robert Haas | 2012-04-24 13:24:58 | Re: Postgres install program files folder |
Previous Message | Robert Haas | 2012-04-24 13:21:38 | Re: DOMAINs and CASTs |