public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
* suggested vhost link speed settings
@ 2010-11-28  2:26 linux_kvm
  2010-11-28 13:03 ` Michael Tokarev
  0 siblings, 1 reply; 2+ messages in thread
From: linux_kvm @ 2010-11-28  2:26 UTC (permalink / raw)
  To: KVM mailing list

Hi list,

Being that the virtio interfaces are stated as acheiving 5-8 Gb
throughput now with vhost, as opposed to 1Gb without, how should their
link speed be defined when the choices are 2500M or 10000M?

I have them plotted out to make a 10Gb bond out of a pair, counting on
5Gb max each, which I'm imagining can be acheived without concern based
on what I read.

If I set them to 'auto-negotiate' will it internally flap between &
cause undesirable consequences?
I don't want to set them at 2500 in case they ever do need to reach
closer to 5 each.

I'm not at a point where I can test anything yet, I'm just planning &
preconfiguring so far.

For the sake of preventing issues down the line I wanted to see if
there's a consensus or standard for this scenario and be as sure as
possible ahead of time.

Thanks,

-C

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: suggested vhost link speed settings
  2010-11-28  2:26 suggested vhost link speed settings linux_kvm
@ 2010-11-28 13:03 ` Michael Tokarev
  0 siblings, 0 replies; 2+ messages in thread
From: Michael Tokarev @ 2010-11-28 13:03 UTC (permalink / raw)
  To: linux_kvm; +Cc: KVM mailing list

28.11.2010 05:26, linux_kvm@proinbox.com wrote:
> Hi list,
> 
> Being that the virtio interfaces are stated as acheiving 5-8 Gb
> throughput now with vhost, as opposed to 1Gb without, how should their
> link speed be defined when the choices are 2500M or 10000M?

The reported link speed for _any_ virtual NIC has no meaning at all,
except of the numbers somewhere in ifconfig or equivalent.  The actual
speed is always limited by the host CPU and the effectiveness of the
implementation currently in use.  So, it is merely a cosmetic thing,
nothing more.

The only possible usage for this number (reported link speed) is to
let some visualisation tools to draw their graphs, to show "link
utilisation" percentage for example.  But for that, there's no actual
utilisation either - it depends on the current host CPU speed, - f.e.
you put additional load to the CPU and the max speed will be lower,
but your visualisator will report low utilization.

In short: don't try to visualize "link utilization", it has no meaning.

As of actual speeds, I can reach ~2Gbps on my somewhat-old-now 2.4GHz
Phenom 9750 (first gen) even without vhost-net.

/mjt

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2010-11-28 13:03 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-11-28  2:26 suggested vhost link speed settings linux_kvm
2010-11-28 13:03 ` Michael Tokarev

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox