From: "Richard Croucher" <richard-jNDFPZUTrfRJpuwtbJ71GdBPR1lH4CV8@public.gmane.org>
To: 'Jabe' <jabe.chapman-9AbUPqfR1/2XDw4h08c5KA@public.gmane.org>,
richard.croucher-jNDFPZUTrfRJpuwtbJ71GdBPR1lH4CV8@public.gmane.org
Cc: 'Ali Ayoub'
<ali-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>,
'Christoph Lameter' <cl-vYTEC60ixJUAvxtiuMwx3w@public.gmane.org>,
'linux-rdma' <linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>,
'sebastien dugue' <sebastien.dugue-6ktuUTfB/bM@public.gmane.org>,
'OF EWG' <ewg-ZwoEplunGu1OwGhvXhtEPSCwEArCW2h5@public.gmane.org>
Subject: RE: [ewg] IPoIB to Ethernet routing performance
Date: Thu, 30 Dec 2010 17:37:51 -0000 [thread overview]
Message-ID: <004301cba848$50253b00$f06fb100$@com> (raw)
In-Reply-To: <4D187DB1.5020005-9AbUPqfR1/2XDw4h08c5KA@public.gmane.org>
IPoIB is far easier to use and does not carry out the additional management
burden of vNICS.
With vNICs you have to manage the MAC address mapping to Ethernet g/w port.
In some situations, such as when multiple G/w's are used for resiliency this
can amount to a lot of separate vNICs on each server to manage. In a small
configuration I had, we ended up with 6 vNICS per server to manage. On a
large configuration this additional management would be a big burden.
My experience with IPoIB has always been very positive. All my existing
socket programs have worked, even some esoteric ioctls I use for multicast
and buffer management.
Performance could always be better, but in my experience it's not great for
the vNICS either. Latency in particular was very disappointing when I
tested.
If you want high performance you have to avoid TCP/IP.
-----Original Message-----
From: Jabe [mailto:jabe.chapman-9AbUPqfR1/2XDw4h08c5KA@public.gmane.org]
Sent: 27 December 2010 11:51
To: richard.croucher-jNDFPZUTrfRJpuwtbJ71GdBPR1lH4CV8@public.gmane.org
Cc: Richard Croucher; 'Ali Ayoub'; 'Christoph Lameter'; 'linux-rdma';
'sebastien dugue'; 'OF EWG'
Subject: Re: [ewg] IPoIB to Ethernet routing performance
On 12/26/2010 11:57 AM, Richard Croucher wrote:
> The vNIC driver only works when you have Ethernet/InfiniBand hardware
> gateways in your environment. It is useful when you have external hosts
to
> communicate with which do not have direct InfiniBand connectivity.
> IPoIB is still heavily used in these environments to provide TCP/IP
> connectivity within the InfiniBand fabric.
> The primary Use Case for vNICs is probably for virtualization servers, so
> that individual Guests can be presented with a virtual Ethernet NIC and do
> not lead to load any InfiniBand drivers. Only the hypervisor needs to
have
> the InfiniBand software stack loaded.
> I've also applied vNICs in the Financial Services arena, for connectivity
to
> external TCP/IP services but there the IPoIB gateway function is arguably
> more useful.
>
> The whole vNIC arena is complicated by different, incompatible
> implementations from each of Qlogic and Mellanox.
>
> Richard
>
Richard, with your explanation I understand why vNIC / EoIB is used in
the case you cite, but I don't understand why it is NOT used in the
other cases (like Ali says).
I can *guess* it's probably because with a virtual ethernet fabric you
have to do all IP stack in software, probably without even having the
stateless offloads (so it would be a performance reason). Is that the
reason?
Thank you
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2010-12-30 17:37 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-12-06 10:24 IPoIB to Ethernet routing performance sebastien dugue
[not found] ` <20101206112454.76bb85f1-xRPE6/W2vR9iM9LT7/dT9zWMkbuR3peG@public.gmane.org>
2010-12-06 10:49 ` Richard Croucher
2010-12-06 11:40 ` [ewg] " sebastien dugue
[not found] ` <20101206124023.025c2f88-xRPE6/W2vR9iM9LT7/dT9zWMkbuR3peG@public.gmane.org>
2010-12-06 12:08 ` Richard Croucher
2010-12-06 13:05 ` sebastien dugue
[not found] ` <20101206140505.20cfc9e2-xRPE6/W2vR9iM9LT7/dT9zWMkbuR3peG@public.gmane.org>
2010-12-09 23:46 ` Christoph Lameter
[not found] ` <alpine.DEB.2.00.1012091745070.29367-sBS69tsa9Uj/9pzu0YdTqQ@public.gmane.org>
2010-12-26 7:43 ` Ali Ayoub
[not found] ` <AANLkTi=2-FJtJCY0+79wyXRszwrhRwQStDTNjRCjr66F-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2010-12-26 10:57 ` Richard Croucher
2010-12-27 11:51 ` Jabe
[not found] ` <4D187DB1.5020005-9AbUPqfR1/2XDw4h08c5KA@public.gmane.org>
2010-12-30 17:37 ` Richard Croucher [this message]
2010-12-28 0:06 ` Ali Ayoub
[not found] ` <AANLkTi=ausgOuEMRe622dNRzVyhN3pxPWeTAJiJihssW-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2010-12-28 15:30 ` Reeted
[not found] ` <4D1A02A8.1060104-9AbUPqfR1/2XDw4h08c5KA@public.gmane.org>
2010-12-29 15:18 ` Tziporet Koren
2011-01-03 22:33 ` Christoph Lameter
2010-12-07 14:42 ` Or Gerlitz
2010-12-06 20:47 ` Jabe
[not found] ` <4CFD4BEE.5070205-9AbUPqfR1/2XDw4h08c5KA@public.gmane.org>
2010-12-06 21:27 ` Jason Gunthorpe
[not found] ` <20101206212759.GB16788-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2010-12-07 7:39 ` sebastien dugue
[not found] ` <20101207083911.2ab47a59-xRPE6/W2vR9iM9LT7/dT9zWMkbuR3peG@public.gmane.org>
2010-12-07 10:02 ` Or Gerlitz
[not found] ` <4CFE0638.2040105-smomgflXvOZWk0Htik3J/w@public.gmane.org>
2010-12-07 10:27 ` sebastien dugue
[not found] ` <20101207112739.0c95db46-xRPE6/W2vR9iM9LT7/dT9zWMkbuR3peG@public.gmane.org>
2010-12-07 10:33 ` Or Gerlitz
[not found] ` <4CFE0D7C.4020904-smomgflXvOZWk0Htik3J/w@public.gmane.org>
2010-12-07 11:48 ` sebastien dugue
[not found] ` <20101207124805.1fede78f-xRPE6/W2vR9iM9LT7/dT9zWMkbuR3peG@public.gmane.org>
2010-12-07 12:32 ` Or Gerlitz
[not found] ` <4CFE2953.2020807-smomgflXvOZWk0Htik3J/w@public.gmane.org>
2010-12-07 13:19 ` sebastien dugue
2010-12-07 13:01 ` Hiroyuki Sato
[not found] ` <AANLkTik2XaM6qyUUc+uzYCzKL+zKbN5kHz2Kf8So8arC-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2010-12-07 13:25 ` sebastien dugue
2010-12-13 16:02 ` Jabe
[not found] ` <4D06438C.9040307-9AbUPqfR1/2XDw4h08c5KA@public.gmane.org>
2010-12-14 16:35 ` [ewg] " Richard Croucher
2010-12-07 7:36 ` sebastien dugue
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='004301cba848$50253b00$f06fb100$@com' \
--to=richard-jndfpzutrfrjpuwtbj71gdbpr1lh4cv8@public.gmane.org \
--cc=ali-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org \
--cc=cl-vYTEC60ixJUAvxtiuMwx3w@public.gmane.org \
--cc=ewg-ZwoEplunGu1OwGhvXhtEPSCwEArCW2h5@public.gmane.org \
--cc=jabe.chapman-9AbUPqfR1/2XDw4h08c5KA@public.gmane.org \
--cc=linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=richard.croucher-jNDFPZUTrfRJpuwtbJ71GdBPR1lH4CV8@public.gmane.org \
--cc=sebastien.dugue-6ktuUTfB/bM@public.gmane.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox