On 8 October 2014 00:34, Peter Geoghegan <pg@heroku.com> wrote:
> INSERTs see #2 win, and by a wider margin than #1 beat #2 with
> UPDATEs. However, insert.sh is by design very unsympathetic towards
> #1. It uses a serial primary key, so every INSERT uselessly obtains a
> HW lock on the same leaf page for the duration of heap insertion.
> Anyway, the median INSERT TPS numbers is 17,759.53 for #1, and
> 20,441.57 TPS for #2. So you're pretty much seeing the full brunt of
> page heavyweight locking, and it isn't all that bad.
Lets see the results of running a COPY please.
-- Simon Riggs http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training & Services