On Tue, Jun 30, 2015 at 3:35 PM, Josh Berkus <josh@agliodbs.com> wrote:
> This goes out Thursday, so I'm putting it here for final
> comments/corrections. Sorry for the last-minute nature of this, but
> time kind of got away from me.
Sorry that I didn't get to this sooner. I have a concern with this item:
* Large speed improvements in in-memory sorting and hashing
The largest acceleration that I've heard anyone have with the
abbreviated keys sorting patch - a 12X improvement, reported by Corey
Huinker - was actually with an external sort (in-memory sorts were
more effectively accelerated during testing on my laptop, but Corey
used a real server).
I think that the "in-memory" qualifier does not add much information
about the hashing improvements, and misleads about the sorting
improvements. Too late for changing the 9.5 alpha notes now, but
something to remember when the beta comes out.
--
Regards,
Peter Geoghegan