Rick,
I have dual core AMD opteron, the IB card are connected to 4X
mellanox infiniscale 2400, I am running Sles 10. when i used to run
the test without any message size (-m )
The SDP test didn't work for me I am try to figure out why also I had
noproblem in loading sdp library.
I also changed the txqueuelen to 4096 the default was 128,
t1:~ # ifconfig ib0
ib0 Link encap:UNSPEC HWaddr 80-00-04-04-
FE-80-00-00-00-00-00-00-00-00-00-00
inet addr:192.168.2.101 Bcast:192.168.2.255 Mask:
255.255.255.0
inet6 addr: fe80::202:c902:24:2909/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:65520 Metric:1
RX packets:11673701 errors:0 dropped:0 overruns:0 frame:0
TX packets:14069560 errors:0 dropped:5 overruns:0 carrier:0
collisions:0 txqueuelen:4096
RX bytes:59793476658 (57023.5 Mb) TX bytes:192547529035
(183627.6 Mb)
t1:~ # netperf -H 192.168.2.100 -c -C --
TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
192.168.2.100 (192.168.2.100) port 0 AF_INET
Recv Send Send Utilization
Service Demand
Socket Socket Message Elapsed Send Recv
Send Recv
Size Size Size Time Throughput local remote
local remote
bytes bytes bytes secs. 10^6bits/s % S % S us/
KB us/KB
87380 16384 16384 10.00 3989.07 12.81 23.66 0.526
0.972
adding the -m with 128K show a better results
netperf -H 192.168.2.100 -c -C -- -m 128K
bytes bytes bytes secs. 10^6bits/s % S % S us/
KB us/KB
87380 16384 131072 10.00 4646.27 12.91 24.76 0.455
0.873
in case of -m 1m the cpu util showed a real different in this case.
bytes bytes bytes secs. 10^6bits/s % S % S us/
KB us/KB
87380 16384 1048576 10.00 5796.64 21.91 31.91
0.619 0.902
Thanks
H. N. Harake
On 11-Sep-2007, at 19:17, Rick Jones wrote:
SDP
_______________________________________________
general mailing list
[email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
