From: Bill Fink <billfink@mindspring.com>
To: Rick Jones <rick.jones2@hp.com>
Cc: Larry McVoy <lm@bitmover.com>,
Linus Torvalds <torvalds@linux-foundation.org>,
davem@davemloft.net, wscott@bitmover.com, netdev@vger.kernel.org
Subject: Re: tcp bw in 2.6
Date: Wed, 3 Oct 2007 03:19:06 -0400 [thread overview]
Message-ID: <20071003031906.5f0d7cfd.billfink@mindspring.com> (raw)
In-Reply-To: <47027C63.803@hp.com>
Tangential aside:
On Tue, 02 Oct 2007, Rick Jones wrote:
> *) depending on the quantity of CPU around, and the type of test one is running,
> results can be better/worse depending on the CPU to which you bind the
> application. Latency tends to be best when running on the same core as takes
> interrupts from the NIC, bulk transfer can be better when running on a different
> core, although generally better when a different core on the same chip. These
> days the throughput stuff is more easily seen on 10G, but the netperf service
> demand changes are still visible on 1G.
Interesting. I was going to say that I've generally had the opposite
experience when it comes to bulk data transfers, which is what I would
expect due to CPU caching effects, but that perhaps it's motherboard/NIC/
driver dependent. But in testing I just did I discovered it's even
MTU dependent (most of my normal testing is always with 9000-byte
jumbo frames).
With Myricom 10-GigE NICs, NIC interrupts on CPU 0 and nuttcp app
running on CPU 1 (both transmit and receive sides), and using 9000-byte
jumbo frames:
[root@lang2 ~]# nuttcp -w10m 192.168.88.16
10078.5000 MB / 10.02 sec = 8437.5396 Mbps 100 %TX 99 %RX
With Myricom 10-GigE NICs, and both NIC interrupts and nuttcp app
on CPU 0 (both transmit and receive sides), again using 9000-byte
jumbo frames:
[root@lang2 ~]# nuttcp -w10m 192.168.88.16
11817.8750 MB / 10.00 sec = 9909.7537 Mbps 100 %TX 74 %RX
Same tests repeated with standard 1500-byte Ethernet MTU:
With Myricom 10-GigE NICs, NIC interrupts on CPU 0 and nuttcp app
running on CPU 1 (both transmit and receive sides), and using
standard 1500-byte Ethernet MTU:
[root@lang2 ~]# nuttcp -M1460 -w10m 192.168.88.16
5685.9375 MB / 10.00 sec = 4768.0951 Mbps 99 %TX 98 %RX
With Myricom 10-GigE NICs, and both NIC interrupts and nuttcp app
on CPU 0 (both transmit and receive sides), again using standard
1500-byte Ethernet MTU:
[root@lang2 ~]# nuttcp -M1460 -w10m 192.168.88.16
4974.0625 MB / 10.03 sec = 4161.6015 Mbps 100 %TX 100 %RX
Now back to your regularly scheduled programming. :-)
-Bill
next prev parent reply other threads:[~2007-10-03 7:19 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20070929142517.EC6AB5FB21@work.bitmover.com>
[not found] ` <alpine.LFD.0.999.0709290914410.3579@woody.linux-foundation.org>
[not found] ` <20070929172639.GB7037@bitmover.com>
[not found] ` <alpine.LFD.0.999.0709291050200.3579@woody.linux-foundation.org>
2007-10-02 0:59 ` tcp bw in 2.6 Larry McVoy
2007-10-02 2:14 ` Linus Torvalds
2007-10-02 2:20 ` Larry McVoy
2007-10-02 3:50 ` David Miller
2007-10-02 4:23 ` Larry McVoy
2007-10-02 15:06 ` John Heffner
2007-10-02 17:14 ` Rick Jones
2007-10-02 17:20 ` Larry McVoy
2007-10-02 18:01 ` Rick Jones
2007-10-02 18:40 ` Larry McVoy
2007-10-02 19:47 ` Rick Jones
2007-10-02 21:32 ` David Miller
2007-10-03 7:19 ` Bill Fink [this message]
2007-10-02 10:52 ` Herbert Xu
2007-10-02 15:09 ` Larry McVoy
2007-10-02 15:41 ` Larry McVoy
2007-10-02 16:25 ` Larry McVoy
2007-10-02 16:47 ` Stephen Hemminger
2007-10-02 16:49 ` Larry McVoy
2007-10-02 17:10 ` Stephen Hemminger
2007-10-15 12:40 ` Daniel Schaffrath
2007-10-15 15:49 ` Stephen Hemminger
2007-10-02 16:34 ` Linus Torvalds
2007-10-02 16:48 ` Larry McVoy
2007-10-02 21:16 ` David Miller
2007-10-02 21:26 ` Larry McVoy
2007-10-02 21:47 ` David Miller
2007-10-02 22:17 ` Rick Jones
2007-10-02 22:32 ` David Miller
2007-10-02 22:36 ` Larry McVoy
2007-10-02 22:59 ` Rick Jones
2007-10-03 8:02 ` David Miller
2007-10-02 16:48 ` Ben Greear
2007-10-02 17:11 ` Larry McVoy
2007-10-02 17:18 ` Ben Greear
2007-10-02 17:21 ` Larry McVoy
2007-10-02 17:54 ` Stephen Hemminger
2007-10-02 18:35 ` Larry McVoy
2007-10-02 18:29 ` John Heffner
2007-10-02 19:07 ` Larry McVoy
2007-10-02 19:29 ` Linus Torvalds
2007-10-02 20:31 ` David Miller
2007-10-02 19:33 ` Larry McVoy
2007-10-02 19:53 ` John Heffner
2007-10-02 20:14 ` Larry McVoy
2007-10-02 20:40 ` Rick Jones
2007-10-02 20:42 ` Wayne Scott
2007-10-02 21:56 ` Linus Torvalds
2007-10-02 19:27 ` Linus Torvalds
2007-10-02 19:53 ` Rick Jones
2007-10-02 20:33 ` David Miller
2007-10-02 20:44 ` Roland Dreier
2007-10-02 21:21 ` Larry McVoy
2007-10-03 21:13 ` Pekka Pietikainen
2007-10-03 21:23 ` Larry McVoy
2007-10-03 21:50 ` Pekka Pietikainen
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=20071003031906.5f0d7cfd.billfink@mindspring.com \
--to=billfink@mindspring.com \
--cc=davem@davemloft.net \
--cc=lm@bitmover.com \
--cc=netdev@vger.kernel.org \
--cc=rick.jones2@hp.com \
--cc=torvalds@linux-foundation.org \
--cc=wscott@bitmover.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.