[176633] in North American Network Operators' Group
RE: 10Gb iPerf kit?
daemon@ATHENA.MIT.EDU (Teleric Team)
Sun Dec 7 11:48:55 2014
X-Original-To: nanog@nanog.org
From: Teleric Team <teleric-lists@outlook.com>
To: Pete Mundy <pete@fiberphone.co.nz>, NANOG <nanog@nanog.org>
Date: Sun, 7 Dec 2014 11:48:47 -0500
In-Reply-To: <70853923-F3DD-4BB7-ABA3-84BB29B3A455@fiberphone.co.nz>
Errors-To: nanog-bounces@nanog.org
> From: pete@fiberphone.co.nz
> Subject: Re: 10Gb iPerf kit?
> Date: Sun=2C 7 Dec 2014 09:24:41 +1300
> To: nanog@nanog.org
>=20
> On 11/11/2014=2C at 1:35 PM=2C Randy Carpenter <rcarpen@network1.net> wro=
te:
>=20
> > I have not tried doing that myself=2C but the only thing that would eve=
n be possible that I know of is thunderbolt.
> > A new MacBook Pro and one of these maybe: http://www.sonnettech.com/pro=
duct/echoexpresssel_10gbeadapter.html
>=20
> Or one of these ones for dual-10Gbit links (one for out of band managemen=
t or internet?):
>=20
> http://www.sonnettech.com/product/twin10g.html
>=20
> I haven't tried one myself=2C but they're relatively cheap (for 10gig) so=
not that much outlay to grab one and try it (esp if you already have an Ap=
ple laptop you can test with).
>=20
How would you use it? with iperf still?I don't think you will go nearly clo=
se to 14.8Mpps per port this way.Unless you are talking about bandwidth tes=
ting with full sized packet frames and low pps rate.
I personally tested a 1Gbit/s port over a MBP retina 15 thunderbot gbe with=
BCM5701 chipset. I had only 220kpps on a single TX flow.Later I tried anot=
her adapter with a marvel yukon mini port. Had better pps rate=2C but nothi=
ng beyond 260kpps.
> I've done loads of 1Gbit testing using the entry-level MacBook Air and a =
Thunderbolt Gigabit Ethernet adapter though=2C and I disagree with Saku's s=
tatement of 'You cannot use UDPSocket like iperf does=2C it just does not w=
ork=2C you are lucky if you reliably test 1Gbps'. I find iperf testing at 1=
Gbit on Mac Air with Thunderbolt Eth extremely reliable (always 950+mbit/se=
c TCP on a good network=2C and easy to push right to the 1gbit limit with U=
DP.
Again=2C with 64byte packet size? Or are you talking MTU?
With MTU size you can try whatever you want and it will seem to be reliable=
. A wget/ftp download of a 1GB file will provide similar results=2C but I d=
ont think this is useful anyway since it won't test anything close to rfc25=
44 or at least an ordinary internet traffic profile with a mix of 600bytes =
pkg size combined with a lower rate of smaller packets (icmp/udp=2C ping/dn=
s/ntp/voice/video).
I am also interested in a cheap and reliable method to test 10GbE connectio=
ns. So far I haven't found something I trust.
>=20
> Pete
>=20
=
