Re: Native partitioning tablespace inheritance - Mailing list pgsql-hackers

From Jonathan S. Katz
Subject Re: Native partitioning tablespace inheritance
Date
Msg-id CADA5465-60CB-4CBC-9A54-38E346273924@excoventures.com
Whole thread Raw
In response to Re: Native partitioning tablespace inheritance  (Christophe Pettus <xof@thebuild.com>)
Responses Re: Native partitioning tablespace inheritance  (Robert Haas <robertmhaas@gmail.com>)
List pgsql-hackers
> On Apr 12, 2018, at 2:36 PM, Christophe Pettus <xof@thebuild.com> wrote:
>
>
>> On Apr 12, 2018, at 09:17, Robert Haas <robertmhaas@gmail.com> wrote:
>> Hmm, that's interesting.  So you want the children to inherit the
>> parent's tablespace when they are created, but if the parent's
>> tablespace is later changed, the existing children don't move?
>
> +1 to that behavior.
>
> While it's always possible to just say "do the right thing" to the application when creating new children (that is,
expectthat they will always specify a tablespace if it's not the default), this seems like the least-surprising
behavior.
>
> It's true that an unpartitioned table will always be created in the default tablespace unless otherwise specified,
butchild tables are sufficiently distinct from that case that I don't see it as a painful asymmetry. 

If there are no strong objections I am going to add this to the “Older Bugs”
section of Open Items in a little bit.

Jonathan



pgsql-hackers by date:

Previous
From: Andrew Gierth
Date:
Subject: Re: submake-errcodes
Next
From: Peter Geoghegan
Date:
Subject: wal_consistency_checking reports an inconsistency on master branch