Re: [CORE] [BUGS] BUG #13350: blindly fsyncing data dir considered harmful - Mailing list pgsql-hackers

From Stephen Frost
Subject Re: [CORE] [BUGS] BUG #13350: blindly fsyncing data dir considered harmful
Date
Msg-id 20150525160626.GO26667@tamriel.snowman.net
Whole thread Raw
In response to Re: [CORE] [BUGS] BUG #13350: blindly fsyncing data dir considered harmful  (Andres Freund <andres@anarazel.de>)
List pgsql-hackers
* Andres Freund (andres@anarazel.de) wrote:
> On May 25, 2015 8:52:33 AM PDT, Bruce Momjian <bruce@momjian.us> wrote:
> >On Mon, May 25, 2015 at 04:37:59PM +0100, Greg Stark wrote:
> >> What exactly is failing?
> >>
> >> Is it that fsync is returning -1 ? Should we just ignore errors from
> >> fsync if it happens in this stage? That might be safer than
> >> determining which files should or shouldn't be fsynced.
> >
> >Interesting idea.  We could skip fsync -1 failures only for symbolic
> >links.
>
> Now there's already a thread on jackets about this.

Agreed- we should probably keep the discussion about fixing it (or
addressing the original issue in some way that doesn't run into this
issue) on the other thread.  This is simply to point out the user impact
and that we need to fix it quickly.
Thanks!
    Stephen

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [CORE] [BUGS] BUG #13350: blindly fsyncing data dir considered harmful
Next
From: Andres Freund
Date:
Subject: Re: fsync-pgdata-on-recovery tries to write to more files than previously