Re: reload-through-the-top-parent switch the partition table

From: Robert Haas <robertmhaas(at)gmail(dot)com>
To: Rushabh Lathia <rushabh(dot)lathia(at)gmail(dot)com>
Cc: PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: reload-through-the-top-parent switch the partition table
Date: 2017-08-02 14:11:46
Message-ID: CA+TgmobsaUcSZbsOBqjOGjXprYrL8cSw0KVg=pfEGkZ1cT3raQ@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Wed, Aug 2, 2017 at 1:01 AM, Rushabh Lathia <rushabh(dot)lathia(at)gmail(dot)com> wrote:
> Looking at the dbObjectTypePriority comments that seems like data
> restoration
> will *absolutely always* follow all CREATE TABLE commands.

Hmm. I wasn't very convinced by those comments, but Tom's commit
a1ef01fe163b304760088e3e30eb22036910a495 convinces me that it has to
work that way. So I think we are OK on that score.

The patch itself looks just fine on a quick glance, modulo the lack of
documentation, but I think we need to bikeshed the name of the flag.
--reload-through-root is clear as daylight to me, but I'm not sure
users will agree. The lack of the word "partition" is perhaps a
significant flaw, and pg_dump doesn't really reload anything; it just
dumps.

The best thing I can come up with after brief thought is
--partition-data-via-root, but maybe somebody else has a better idea?

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2017-08-02 14:17:12 Re: pgbench: Skipping the creating primary keys after initialization
Previous Message Robert Haas 2017-08-02 14:04:23 Re: pgbench: Skipping the creating primary keys after initialization