From: Tom Lane <tgl@sss.pgh.pa.us>
To: Tobias Bussmann <t.bussmann@gmx.net>
cc: Mark Guertin <markguertin@gmail.com>, pgsql-bugs@lists.postgresql.org
Subject: Re: pg_basebackup: errors on macOS on directories with ".DS_Store"
 files
In-reply-to: <794684A8-D3FF-4109-815F-C9EB66AF65EA@gmx.net>
References: <E258CE50-AB0E-455D-8AAD-BB4FE8F882FB@gmail.com>
 <794684A8-D3FF-4109-815F-C9EB66AF65EA@gmx.net>
Comments: In-reply-to Tobias Bussmann <t.bussmann@gmx.net>
	message dated "Wed, 19 Apr 2023 19:51:55 +0200"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <3000520.1681928151.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Wed, 19 Apr 2023 14:15:51 -0400
Message-ID: <3000521.1681928151@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/3000521.1681928151%40sss.pgh.pa.us>
Precedence: bulk

Tobias Bussmann <t.bussmann@gmx.net> writes:
>> pg_basebackup: error: backup failed: ERROR:  invalid segment number 0 i=
n file ".DS_Store"
> This seem to be the same problem as already discussed in [1]: Tools like=
 pg_basebackup and pg_checksums can get upset if there are random extra fi=
les in $PGDATA/base, /global or /pg_tblspc. Unfortunately, macOS has the h=
abit to create such files quite easily. If I read the outcome correctly, w=
e are back at a blacklist approach, so adding ".DS_Store" may be something=
 to consider. =

> [1]: <https://www.postgresql.org/message-id/flat/20181021134206.GA14282%=
40paquier.xyz>

Yeah.  I wonder if we ought to do something more general, and
ignore all files whose names start with ".".

			regards, tom lane