From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Reese Faucette" Subject: Re: vif interface dropping 80% packets Date: Tue, 6 Feb 2007 10:58:37 -0800 Message-ID: <007101c74a20$cf49d600$58c31fac@bart> References: <086c01c74986$0e8fd120$58c31fac@bart> <08CA2245AFCF444DB3AC415E47CC40AF71781F@G3W0072.americas.hpqcorp.net> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============0732765472==" Return-path: List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Mime-version: 1.0 Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: xen-devel@lists.xensource.com List-Id: xen-devel@lists.xenproject.org This is a multi-part message in MIME format. --===============0732765472== Content-Type: multipart/alternative; boundary="----=_NextPart_000_006E_01C749DD.C0DD30F0" This is a multi-part message in MIME format. ------=_NextPart_000_006E_01C749DD.C0DD30F0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Most likely your dom0 CPU is saturated.. It is unrealistic to expect = full throughput of a 10 Gig NIC. I would be surprised if even linux = could keep up with a receive rate of 10 Gb/s. Did you try the same = experiment on native linux? In my experiments, on a 4-way 2.8 Ghz Xeon, = dom0 consumes ~75% of one cpu for processing receive packets while the = guest consume ~55% of another CPU. With native->native or even native->dom0, I get 9.88 Gb/s with 21% = utilization of receive side CPU. It's a nice machine and we make nice = NICs ;-) So, it's not unrealistic. It is expected that your dual CPU system saturates at a rate slighlty = above 1 Ghz. (You are also consuming extra cycles to drop packets) You could confirm that the CPU is saturating by running xentop while = running your experiments. You could avoid dropping packets and get a more realistic experiment = by running a TCP experiment as Ian suggested i have a pending note on that, TCP test is actually what I tried first, = and the perf was terrible due to the same packet loss. more info to = come. -reese ------=_NextPart_000_006E_01C749DD.C0DD30F0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

 
Most likely your dom0 CPU is saturated.. It = is=20 unrealistic to expect full throughput of a 10 Gig NIC. I would be = surprised if=20 even linux could keep up with a receive rate of 10 = Gb/s. Did=20 you try the same experiment on native linux? In  my experiments, = on a=20 4-way 2.8 Ghz Xeon, dom0 consumes ~75% of one cpu for processing = receive=20 packets while the guest consume ~55% of another=20 CPU.
With native->native or even native->dom0, = I get 9.88=20 Gb/s with 21% utilization of receive side CPU.  It's a nice machine = and we=20 make nice NICs ;-)  So, it's not unrealistic.
It is  expected that your dual = CPU system=20 saturates at a rate slighlty above 1 Ghz. (You are also consuming = extra cycles=20 to drop packets)
You could confirm that = the CPU is=20 saturating by running xentop while running your=20 experiments.
You could avoid dropping = packets and get=20 a more realistic experiment by running a TCP experiment as Ian=20 suggested
i have a pending note on that, TCP test is = actually=20 what I tried first, and the perf was terrible due to the same packet = loss. =20 more info to come.
-reese
 
------=_NextPart_000_006E_01C749DD.C0DD30F0-- --===============0732765472== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel --===============0732765472==--