Hi Miles,
Results below
ke...@opensolaris-ws-01:~$ kstat -m e1000g |grep Reset
Reset Count 0
ke...@opensolaris-ws-01:~$ iperf -c 192.168.0.240 -f M
------------------------------------------------------------
Client connecting to 192.168.0.240, TCP port 5001
TCP window size: 0.05 MByte (default)
------------------------------------------------------------
[ 3] local 192.168.0.250 port 37249 connected with 192.168.0.240 port 5001
[ ID] Interval Transfer Bandwidth
[ 3] 0.0-10.0 sec 477 MBytes 47.6 MBytes/sec
ke...@opensolaris-ws-01:~$ kstat -m e1000g |grep Reset
Reset Count 0
Thanks
Kevin
----- Original Message -----
From: "Min Miles Xu" <min...@sun.com>
To: "Kevin Martin" <kevin.mart...@free.fr>
Cc: networking-discuss@opensolaris.org
Sent: Monday, 28 December, 2009 06:21:47 GMT +00:00 GMT Britain, Ireland,
Portugal
Subject: Re: [networking-discuss] Significant performance diff between e1000g0
and bge0
Hi Kevin,
Could you show me the "kstat -m e1000g |grep Reset" output before and
after the iperf test?
Thanks,
Miles
Kevin Martin wrote:
> On svn_130 and svn_111b running on 2x hp ml115 g5 with quad opterons with two
> network interfaces. Broadcom BCM 5722 (pci-e) and intel pro 1000 (82541 pci)
> connected to the same 3com 2924 gigabit switch
> When I run iperf on bge0 I get 113 MegaBytes/s, on e1000g0 I get 54
> MegaBytes/s. I've changed cables, ports on the switch. Any idea's why there
> is such a large difference in the performance?
>
> Thanks Kevin
>
_______________________________________________
networking-discuss mailing list
networking-discuss@opensolaris.org
