Message-ID: <5539592F.1020909@pgmasters.net>
Date: Thu, 23 Apr 2015 16:42:23 -0400
From: David Steele <david@pgmasters.net>
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.10;
 rv:31.0) Gecko/20100101 Thunderbird/31.6.0
MIME-Version: 1.0
To: Andres Freund <andres@anarazel.de>,
 Robert Haas <robertmhaas@gmail.com>
CC: Bruce Momjian <bruce@momjian.us>, Josh Berkus <josh@agliodbs.com>,
 Stephen Frost <sfrost@snowman.net>,
 "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Subject: Re: tablespaces inside $PGDATA considered harmful
References: 
 <WM!43fd2162c8c08343aade3fbea02e9c2844064c7f40cf603f5080ab78b7a3fbf2728d833090982f0743f5367f06e5a45b!@asav-2.01.com>
 <54CBD519.8030502@agliodbs.com>
 <WM!eb076638a020a207a4a7725ea9d3d99d976a0bfbbd2697fb45c5ea8dd966ae07a6f89f0fffaf4a7cd23e49aaf7416845!@asav-3.01.com>
 <54CBF6FE.8030209@agliodbs.com> <20150423024102.GJ13362@momjian.us>
 <CA+TgmoZiCt5=4TRvLofhKWLkxphgsyPAGWueFn1bSgt+m9CzKQ@mail.gmail.com>
 <20150423150043.GA17899@momjian.us>
 <CA+TgmoaEqZJAqPqa-snZWTy9=cOui5AJgaVDBpBDLk2+Za-KAw@mail.gmail.com>
 <20150423195742.GK3055@alap3.anarazel.de>
 <CA+Tgmob+5jCELs_mb=fFHBHJJhOaE=gTkU-6Cr0rvcUrvXpnDQ@mail.gmail.com>
 <20150423203040.GN3055@alap3.anarazel.de>
In-Reply-To: <20150423203040.GN3055@alap3.anarazel.de>
Content-Type: multipart/signed; micalg=pgp-sha256;
 protocol="application/pgp-signature";
 boundary="oha7USfiHwmL6gd13DwaBlsbFp0M3LGWN"
Precedence: bulk
Sender: pgsql-hackers-owner@postgresql.org

This is an OpenPGP/MIME signed message (RFC 4880 and 3156)
--oha7USfiHwmL6gd13DwaBlsbFp0M3LGWN
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

On 4/23/15 4:30 PM, Andres Freund wrote:
> On 2015-04-23 16:26:09 -0400, Robert Haas wrote:
>> But pg_upgrade automates all that, so you can't use pg_upgrade in that=

>> case.  If we add a GUC as I suggested, you can still use pg_upgrade.
>=20
> But we also have to live with data directories being in a shit state
> forever onward. We won't really be able to remove the option
> realistically.
>=20
> It's not that hard to just move the tablespace out of the data director=
y
> while the server. As long as you move it on the same partition, it's
> even fast.

I agree.  It wouldn't be that hard to do a bit of directory manipulation
before upgrading - and that's only for the people who have put
tablespaces in $PGDATA.  I've never seen it before, but I have no doubt
that it happens.  I can see how it might make a weird sort of sense
depending on the level of experience.

--=20
- David Steele
david@pgmasters.net


--oha7USfiHwmL6gd13DwaBlsbFp0M3LGWN
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
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=mJWm
-----END PGP SIGNATURE-----

--oha7USfiHwmL6gd13DwaBlsbFp0M3LGWN--