On Wed, Mar 07, 2001 at 09:41:35PM +1100, Darren Reed wrote:
> For a web server, the appropriate system call to benchmark is accept,
> not connect, unless I'm mistaken ?  Is hbench using two different
> machines to measure this ?  Otherwise, how do know if you're doing
> a connect from A to A which half the delay is really in ?

Charles already pointed out that the calls to inpcb_lookup_port
come from inpcbbind, often due to the implicit bind that is
done when doing a connect(). So yes, my webserver comment doesn't
really apply. However, this still remains a waste of CPU cycles,
and should be improved.

- Frank

-- 
Frank van der Linden                           fvdl@wasabisystems.com
======================================================================
Quality NetBSD CDs, Support & Service.   http://www.wasabisystems.com/