Jan Nielsen <jan(dot)sture(dot)nielsen(at)gmail(dot)com> wrote:
> The range 2048-8192 is an area to study in more detail, obviously.
> Feedback welcome.
I don't see what's to study there, really. Performance drops off
when database size grows from 30GB to 60GB on a system with 48GB
RAM. And even more when you double database size again. Access to
disk is slower than access to system RAM. Is there something else I
should notice that I'm missing?
The local dips in the list suggest that you're not controlling for
checkpoints or autovacuum runs as well as you might, or that you're
not using a large enough number of runs at each scale.
-Kevin