On Tue, Jan 10, 2012 at 3:16 AM, Simon Riggs <simon@2ndquadrant.com> wrote:
> So benchmarking write-heavy workloads and separately benchmarking
> read-only workloads is more representative.
Absolutely. High write activity applications are much more difficult
to optimize with simple tricks like client side caching. Also,
storage is finally moving out of the dark ages so that high write
transaction rate servers are no longer necessarily i/o bound without
on reasonable hardware.
merlin