Re: Potential data loss of 2PC files

From: Michael Paquier <michael(dot)paquier(at)gmail(dot)com>
To: Teodor Sigaev <teodor(at)sigaev(dot)ru>
Cc: Heikki Linnakangas <hlinnaka(at)iki(dot)fi>, Andres Freund <andres(at)anarazel(dot)de>, Robert Haas <robertmhaas(at)gmail(dot)com>, PostgreSQL mailing lists <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Potential data loss of 2PC files
Date: 2017-03-24 23:15:18
Message-ID: CAB7nPqS0+=OBe_w9FGtdmrqdt6RPWUJMJr_TYXH00UaBxfcoPg@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Fri, Mar 24, 2017 at 11:36 PM, Teodor Sigaev <teodor(at)sigaev(dot)ru> wrote:
>> And the renaming of pg_clog to pg_xact is also my fault. Attached is
>> an updated patch.
>
>
> Thank you. One more question: what about symlinks? If DBA moves, for
> example, pg_xact to another dist and leaves the symlink in data directoty.
> Suppose, fsync on symlink will do nothing actually.

I did not think of this case, but is that really common? There is even
no option to configure that at command level. And surely we cannot
support any fancy scenario that people figure out using PGDATA.
Existing callers of fsync_fname don't bother about this case as well
by the way, take the calls related to pg_logical and pg_repslot.

If something should be done in this area, that would be surely in
fsync_fname directly to centralize all the calls, and I would think of
that as a separate patch, and a separate discussion.
--
Michael

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Alvaro Herrera 2017-03-24 23:27:10 dsm.c API tweak
Previous Message David Steele 2017-03-24 23:13:21 Re: increasing the default WAL segment size