On 2014-10-16 10:56:46 +0100, Simon Riggs wrote:
> On 16 October 2014 10:00, Andres Freund <andres@2ndquadrant.com> wrote:
>
> >> The bug lies in the default behaviour, which we must fix.
> >
> > But the bug is unrelated to manual checkpoints. So I don't think this is
> > really related.
>
> Yes, agreed; I was just thinking that myself.
>
> We need to decouple the bug fix from any behaviour change of the
> CHECKPOINT command, so v3 is not the right thing for backpatching.
I'm not following? v3 is the one that doesn't change the behaviour of a
manual CHECKPOINT? v1/v2 changed the behaviour of it?
Greetings,
Andres Freund
--
Andres Freund http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services