Hi everbody.

Shortly ago i sent a mail because of raid and cgd.
Meanwhile i figured out that i'm having a general problem with high
loads (~4-10). When i'm doing a ping from a machine in the network (they are
connected via a cross cable to be exact) some of the ping response times
really are not what one could expect from an 100mbit cross link.
For example:
jochen@columbus ~> ping 192.168.0.100 | grep -v "time=0."
PING 192.168.0.100 (192.168.0.100): 56 data bytes
64 bytes from 192.168.0.100: icmp_seq=270 ttl=255 time=75.670 ms
64 bytes from 192.168.0.100: icmp_seq=455 ttl=255 time=98.058 ms
64 bytes from 192.168.0.100: icmp_seq=701 ttl=255 time=28.991 ms
64 bytes from 192.168.0.100: icmp_seq=718 ttl=255 time=1.031 ms
64 bytes from 192.168.0.100: icmp_seq=1132 ttl=255 time=1.503 ms
64 bytes from 192.168.0.100: icmp_seq=1353 ttl=255 time=19.023 ms
64 bytes from 192.168.0.100: icmp_seq=1540 ttl=255 time=1.134 ms
64 bytes from 192.168.0.100: icmp_seq=1683 ttl=255 time=60.135 ms

Where 192.168.0.100 is the netbsd machine.
I've never experienced such a behaviour with my FreeBSD machine and the
fact that i really start liking NetBSD will make it even harder for me
to switch back.
I'm using a dual piii machine with NetBSD 2.0 and a SMP kernel.
Dmesg and kernel configuration file have already been sent to this list:
http://mail-index.netbsd.org/netbsd-users/2005/01/03/0009.html

Something is coming to my mind just at the moment.
Maybe it's a problem with one of the NIC's. I will give that a try as
soon as i am at home again (there is an onboard fxp0 which is currently
connected and a wm0 which i'll try out later).
But i suppose that it's a problem with the kernel..

Please help me even your wildest speculations are welcome :)

Best Regards,

Jochen Keil