Tom Lane wrote:
> Bruce Momjian <bruce@momjian.us> writes:
> > Alvaro Herrera wrote:
> >> I thought it was impossible to use bare mountpoints as tablespaces due
> >> to ownership problems ... Is that not the case? -1 for special hacks
> >> that work around bogus setups, if that means intrusive changes to the
> >> core code.
>
> > I talked to the person who reported the problem and he and I confirmed
> > that it is quite easy to make the mount point be owned by the postgres
> > user and have that function as a tablespace. Is that not a supported
> > setup?
>
> It might be *possible*, but that doesn't make it a good idea. The
> traditional sysadmin advice in this area is that mount points should
> be owned by root. I don't really remember the reasoning but I'm pretty
> sure I remember the principle.
Yea, I think the logic is that files under that directory disappear
after the mount, so you don't want users putting things in there
accidentally. In fact, the user said they mount a pg_xlog directory
under the $PGDATA directory (rather than use a symlink), which I also
thought was an odd approach and prone to problems if the mount failed
intermittently.
> > It was actually Tom's idea months ago to put a version-specific
> > directory in the tablespace.
>
> I was just about to re-suggest that. Why do you think it's such a
> bad idea?
I liked the idea, but I listed it as item #2 and no one else said they
liked it. The only complexity I can see with the idea is that doing an
upgrade from one alpha to another would have the same major version
number and would therefore not be possible, so maybe we have to use the
catalog version number in there.
-- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB
http://enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +