netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jesper Dangaard Brouer <brouer@redhat.com>
To: Jesper Dangaard Brouer <brouer@redhat.com>
Cc: "Tom Herbert" <therbert@google.com>,
	"Jamal Hadi Salim" <jhs@mojatatu.com>,
	"Eric Dumazet" <eric.dumazet@gmail.com>,
	"Linux Netdev List" <netdev@vger.kernel.org>,
	"David S. Miller" <davem@davemloft.net>,
	"Alexander Duyck" <alexander.h.duyck@intel.com>,
	"Toke Høiland-Jørgensen" <toke@toke.dk>,
	"Florian Westphal" <fw@strlen.de>,
	"Dave Taht" <dave.taht@gmail.com>,
	"John Fastabend" <john.r.fastabend@intel.com>,
	"Daniel Borkmann" <dborkman@redhat.com>,
	"Hannes Frederic Sowa" <hannes@stressinduktion.org>
Subject: Re: [net-next PATCH 1/1 V4] qdisc: bulk dequeue support for qdiscs with TCQ_F_ONETXQUEUE
Date: Mon, 29 Sep 2014 22:14:31 +0200	[thread overview]
Message-ID: <20140929221431.3e322aab@redhat.com> (raw)
In-Reply-To: <20140925172329.7460f787@redhat.com>


On Thu, 25 Sep 2014 17:23:29 +0200 Jesper Dangaard Brouer <brouer@redhat.com> wrote:
 
> I will redo the tests, once I get home to my testlab, as the remote lab
> I'm using now is annoyingly slow rebooting machines, as we not longer
> have a runtime option for enable/disable (I'm currently in Switzerland).   

I'm going to change the bulking "budget" from 8 packet to 7 packets,
based on my netperf-wrapper test.

I've designed some tests for the tool "netperf-wrapper" that tries to
measure the HoL blocking effect.  I've turned down the link speed to
100Mbit/s on driver igb.  And uses a single TXQ setup
(cmdline: ethtool -L eth1 combined 1).

I can now show some kind of HoL blocking effect.  The strange part is
the HoL effect only occurs above 8 packets, 7 packet and below show no
bad effect of this bulking patch.

Results 100Mbit/s test qdisc_prio_hol, latency in high prio band:
 * GSO: stable average on 2.23ms
 * TSO: varies between min 4.13ms to 4.41ms (range 0.28ms)
 * No bulking: stable average on 1.71ms (3x outliners on 1.95ms)
 * Bulking(8): varies between 1.71ms to 1.95ms (range 0.24ms)
 * Bulking(7): stable average on 1.71ms (1x outliner  on 1.95ms)
 * Bulking(6): stable average on 1.71ms (3x outliners on 1.91ms)
 * Bulking(5): stable average on 1.71ms (1x outliner  on 1.91ms)
 * Bulking(5): stable average on 1.71ms (1x outliner  on 1.91ms)
 * Bulking(4): stable average on 1.71ms (0x outliner)

Bulking(8) the calculation:
 * Added delay were 0.24ms (1.95 - 1.71ms) corrosponding to 3000 bytes
 * 1500 bytes *8 / 100Mbit = 0.12 ms
 * (100*10^6)*(1.95/10^3)/8 = 24375 bytes
 * (100*10^6)*(1.71/10^3)/8 = 21375 bytes
 * Added HoL: 3000 bytes
 * Still compared to TSO and GSO, the added HoL blocking is small.


-- 
Best regards,
  Jesper Dangaard Brouer
  MSc.CS, Sr. Network Kernel Developer at Red Hat
  Author of http://www.iptv-analyzer.org
  LinkedIn: http://www.linkedin.com/in/brouer

  parent reply	other threads:[~2014-09-29 20:15 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-09-24 16:10 [net-next PATCH 0/1 V4] qdisc bulk dequeuing and utilizing delayed tailptr updates Jesper Dangaard Brouer
2014-09-24 16:12 ` [net-next PATCH 1/1 V4] qdisc: bulk dequeue support for qdiscs with TCQ_F_ONETXQUEUE Jesper Dangaard Brouer
2014-09-24 17:23   ` Eric Dumazet
2014-09-24 17:58     ` Jesper Dangaard Brouer
2014-09-24 18:34       ` Tom Herbert
2014-09-24 19:22         ` Eric Dumazet
2014-09-25  2:12           ` Eric Dumazet
2014-09-25  2:38             ` Tom Herbert
2014-09-25  2:58               ` Dave Taht
2014-09-26 18:06                 ` Eric Dumazet
2014-09-25 23:40             ` Eric Dumazet
2014-09-26  6:04               ` [PATCH net-next] dql: dql_queued() should write first to reduce bus transactions Eric Dumazet
2014-09-26  8:47                 ` Jesper Dangaard Brouer
2014-09-26 11:06                 ` Hannes Frederic Sowa
2014-09-26 12:02                   ` Eric Dumazet
2014-09-28 21:43                 ` David Miller
2014-09-26  9:23               ` [net-next PATCH 1/1 V4] qdisc: bulk dequeue support for qdiscs with TCQ_F_ONETXQUEUE David Laight
2014-09-26 13:16                 ` David Laight
2014-09-26 13:38                   ` Eric Dumazet
2014-09-24 22:13       ` Jamal Hadi Salim
2014-09-25  8:25         ` Jesper Dangaard Brouer
2014-09-25 12:48           ` Jamal Hadi Salim
2014-09-25 14:40             ` Tom Herbert
2014-09-25 14:57               ` Jesper Dangaard Brouer
2014-09-25 15:05                 ` Tom Herbert
2014-09-25 15:23                   ` Jesper Dangaard Brouer
2014-09-25 15:58                     ` Tom Herbert
2014-09-29 20:23                       ` Jesper Dangaard Brouer
2014-09-29 20:14                     ` Jesper Dangaard Brouer [this message]
2014-09-25 15:12                 ` Eric Dumazet
2014-09-25 13:52           ` Dave Taht

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=20140929221431.3e322aab@redhat.com \
    --to=brouer@redhat.com \
    --cc=alexander.h.duyck@intel.com \
    --cc=dave.taht@gmail.com \
    --cc=davem@davemloft.net \
    --cc=dborkman@redhat.com \
    --cc=eric.dumazet@gmail.com \
    --cc=fw@strlen.de \
    --cc=hannes@stressinduktion.org \
    --cc=jhs@mojatatu.com \
    --cc=john.r.fastabend@intel.com \
    --cc=netdev@vger.kernel.org \
    --cc=therbert@google.com \
    --cc=toke@toke.dk \
    /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;
as well as URLs for NNTP newsgroup(s).