Re: Compression of full-page-writes

On Fri, Jan 2, 2015 at 11:52 AM, Bruce Momjian <bruce(at)momjian(dot)us> wrote:
> OK, so given your stats, the feature give a 12.5% reduction in I/O. If
> that is significant, shouldn't we see a performance improvement? If we
> don't see a performance improvement, is I/O reduction worthwhile? Is it
> valuable in that it gives non-database applications more I/O to use? Is
> that all?
>
> I suggest we at least document that this feature as mostly useful for
> I/O reduction, and maybe say CPU usage and performance might be
> negatively impacted.
>
> OK, here is the email I remember from Fujii Masao this same thread that
> showed a performance improvement for WAL compression:
>
> http://www.postgresql.org/message-id/CAHGQGwGqG8e9YN0fNCUZqTTT=hNr7Ly516kfT5ffqf4pp1qnHg@mail.gmail.com
>
> Why are we not seeing the 33% compression and 15% performance
> improvement he saw? What am I missing here?

Bruce, some database workloads are I/O bound and others are CPU bound.
Any patch that reduces I/O by using CPU is going to be a win when the
system is I/O bound and a loss when it is CPU bound. I'm not really
sure what else to say about that; it seems pretty obvious.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company