Re: Locking B-tree leafs immediately in exclusive mode

On Fri, Jul 27, 2018 at 6:11 PM Alexander Korotkov
<a(dot)korotkov(at)postgrespro(dot)ru> wrote:
> On Fri, Jul 27, 2018 at 12:30 AM Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
> > On 26 July 2018 at 20:59, Alexander Korotkov <a(dot)korotkov(at)postgrespro(dot)ru> wrote:
> >
> > > Great, thank you! So, I think the regression is demystified. We can
> > > now conclude that on our benchmarks this patch doesn't cause
> > > performance regression larger than measurement error. But in some
> > > cases it shows huge performance benefit.
> > >
> > > So, I'm going to commit this, if no objections.
> >
> > +1 to commit.
> >
> > What will the commit message be?
> >
> > For me, this is about reducing contention on index leaf page hotspots,
> > while at the same time reducing the path length of lock acquisition on
> > leaf pages
>
> So, reducing path length of lock acquisition is particular technical
> change made, while reducing contention on index leaf pages is a
> result. I think that reducing path length of lock acquisition should
> be mentioned in title of commit message, while contention reduction
> should be mentioned in the body of commit message, because it's
> motivation of this commit. If we would have release notes item for
> this commit, it should also mention contention reduction, because it's
> a user-visible effect of this commit.

So, I've pushed it. Thanks to everybody for review.

------
Alexander Korotkov
Postgres Professional: http://www.postgrespro.com
The Russian Postgres Company