>>>>> "CS" == Curt Sampson <cjs@cynic.net> writes:
CS> For most workloads, in the long run, that will force you to do disk
CS> I/O that you would not have had to do otherwise. A single disk I/O
CS> is far more expensive than hundreds of copies between the OS buffer
CS> cache and postgres' shared memory.
It occurs to me that these two opposing factors could be plotted and
make somewhat of a line each. The ideal point would be where they
crossed.
Perhaps one day when I've got a few hours I'll copy over my large
dataset to my devel server and try to mimic the load with different
shm sizes...
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Vivek Khera, Ph.D. Khera Communications, Inc.
Internet: khera@kciLink.com Rockville, MD +1-240-453-8497
AIM: vivekkhera Y!: vivek_khera http://www.khera.org/~vivek/