Netdev List
 help / color / mirror / Atom feed
From: Eric Dumazet <eric.dumazet@gmail.com>
To: starlight@binnacle.cx
Cc: linux-kernel@vger.kernel.org, netdev <netdev@vger.kernel.org>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>
Subject: Re: big picture UDP/IP performance question re 2.6.18 -> 2.6.32
Date: Sat, 01 Oct 2011 21:11:43 +0200	[thread overview]
Message-ID: <1317496303.3802.25.camel@edumazet-laptop> (raw)
In-Reply-To: <6.2.5.6.2.20111001141002.05af4b20@binnacle.cx>

Le samedi 01 octobre 2011 à 14:16 -0400, starlight@binnacle.cx a écrit :
> At 08:44 AM 10/1/2011 +0200, Eric Dumazet wrote:
> >In my experience, I have the exact opposite :
> >performance greatly improved in recent
> >kernels. Unless you compile your kernel to include
> >new features that might reduce performance
> >(namespaces, cgroup, ...)
> 
> RH has both of the above turned on in the
> 2.6.32-71.29.1.el6.x86_64 kernel tested.
> 
> If these are big negatives to network
> performance, could you list what should
> specifically turned off to maximize
> results?  Also a recommendation for
> the best recent kernel for another
> benchmark would be helpful.
> 
> Probably can't convince anyone to deploy a
> kernel without commercial support, but if
> an alternate compile fixes performance it
> might be possible to convince RH to support
> the alternative build.
> 

2.6.32 has a perf tool, that can really help to spot in a few minutes
hot spots. That would definitely help to further diagnose what could be
the problem in your workload.

A single patch can have huge performance impact, sometime not noticed.

For example, in 2.6.36, AF_UNIX support for pid namespaces dropped
performance a lot [commit 7361c36c5224 (af_unix: Allow credentials to
work across user and pid namespaces)], because of a single atomic
operation, but done on each send() and receive() on a central location.

  parent reply	other threads:[~2011-10-01 19:11 UTC|newest]

Thread overview: 47+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-10-01 18:16 big picture UDP/IP performance question re 2.6.18 -> 2.6.32 starlight
2011-10-01 18:40 ` Willy Tarreau
2011-10-01 19:11 ` Eric Dumazet [this message]
2011-10-01 19:43   ` starlight
  -- strict thread matches above, loose matches on Subject: below --
2011-10-07  3:27 starlight
2011-10-07  5:40 ` Eric Dumazet
2011-10-07  6:13   ` starlight
2011-10-07 18:09     ` chetan loke
     [not found]       ` <CAAsGZS4s1wTWW1j7FRUWW9jqpPUVF3Q46AMa7+njvE1ckX0Snw @mail.gmail.com>
2011-10-07 18:37         ` starlight
2011-10-07 19:27           ` chetan loke
     [not found]             ` <CAAsGZS4b2F9N3nV3TNu5xG+=2d0L0ncste4xv2vqoVFb1pOxEw @mail.gmail.com>
2011-10-07 19:41               ` starlight
2011-10-07 20:07           ` Ben Hutchings
2011-10-11 16:24   ` Chris Friesen
2011-10-07  2:33 starlight
2011-10-07  2:24 starlight
2011-10-05  6:58 starlight
2011-10-05  8:53 ` Eric Dumazet
     [not found]   ` <1317804832.2473.25.camel@edumazet-HP-Compaq-6005-Pr o-SFF-PC>
2011-10-05 11:50     ` starlight
2011-10-05  6:11 starlight
2011-10-05  3:35 starlight
2011-10-03 18:02 starlight
2011-10-05  6:53 ` Eric Dumazet
2011-10-03 15:25 starlight
2011-10-03 16:16 ` Eric Dumazet
     [not found]   ` <1317658588.2442.5.camel@edumazet-HP-Compaq-6005-Pro -SFF-PC>
2011-10-03 16:28     ` starlight
2011-10-04 19:16 ` Christoph Lameter
2011-10-04 19:38   ` Joe Perches
2011-10-04 19:42     ` Christoph Lameter
2011-10-04 19:49       ` Serge Belyshev
2011-10-04 20:03         ` Christoph Lameter
2011-10-04 20:12           ` Serge Belyshev
2011-10-04 22:32             ` Con Kolivas
2011-10-04 19:45     ` starlight
2011-10-05 13:22   ` Peter Zijlstra
2011-10-05 14:26     ` Christoph Lameter
2011-10-05 15:12       ` Andi Kleen
2011-10-05 15:33       ` Peter Zijlstra
2011-10-05 15:12     ` starlight
2011-10-02  5:33 starlight
2011-10-02  7:21 ` Eric Dumazet
2011-10-02  8:03   ` Eric Dumazet
2011-10-02 14:47   ` Stephen Hemminger
2011-10-02 15:06   ` starlight
2011-10-04 19:54     ` Loke, Chetan
2011-10-01 21:13 starlight
     [not found] <6.2.5.6.2.20111001012019.05c05b80@flumedata.com>
2011-10-01  6:44 ` Eric Dumazet
2011-10-01 15:56   ` starlight

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=1317496303.3802.25.camel@edumazet-laptop \
    --to=eric.dumazet@gmail.com \
    --cc=a.p.zijlstra@chello.nl \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=starlight@binnacle.cx \
    /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