Ethernet Bridge development
 help / color / mirror / Atom feed
From: <james.card@nokia.com>
To: bridge@lists.linux-foundation.org
Subject: [Bridge] Slow receive side on bridge
Date: Mon, 22 Feb 2010 17:04:43 +0100	[thread overview]
Message-ID: <C7A8154B.1201%james.card@nokia.com> (raw)
In-Reply-To: <2e59e6971002211647r10f229b2l25b5dd68e397dfff@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 2061 bytes --]

Hi,

I have a CentOS 5.4 distro running Xen on top of a Linux bridge (set up by Xen).  I have noticed by using iPerf in bidirectional mode that the traffic going into the server running the bridge is quite a bit slower with the bridge enabled than when it isn't.   The HW is a multiblade setup (8 blades each running CentOS 5.x).  On one of the blades, I shut down Xen and initially run with the bridge shut down and transferred a file from another machine using scp.  I then started up the bridge and transferred the same file and you can see below there is a big difference (3.6MB/s vs 23.2KB/s).  I stopped Xen before doing the following...

Any help on where to look would be helpful.  I was going to shut down all but one blade to see if it could be a lack of STP causing issues.

-- Xen bridge initially shut down

[root@bsbchp003 ~]# brctl show
bridge name    bridge id        STP enabled    interfaces
virbr0        8000.000000000000    yes

-- transfer the 3.7GB file

[root@bsbchp003 ~]# scp card@harvard:bigfile .
bigfile                                                                                                                                                                 100% 3685KB   3.6MB/s   00:00

-- start the bridge

[root@bsbchp003 ~]# /etc/xen/scripts/network-bridge start
net.bridge.bridge-nf-call-arptables = 0
net.bridge.bridge-nf-call-ip6tables = 0
net.bridge.bridge-nf-call-iptables = 0
Nothing to flush.
Nothing to flush.
Waiting for peth0 to negotiate link..[root@bsbchp003 ~]#

-- the Xen bridge is now up

[root@bsbchp003 ~]# brctl show
bridge name    bridge id        STP enabled    interfaces
virbr0        8000.000000000000    yes
xenbr0        8000.feffffffffff    no        peth0
                            vif0.0
[root@bsbchp003 ~]# scp card@harvard:bigfile .
bigfile                                                                                                                                                                 100% 3685KB  23.2KB/s   02:39
[root@bsbchp003 ~]#

[-- Attachment #2: Type: text/html, Size: 4661 bytes --]

      parent reply	other threads:[~2010-02-22 16:04 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-02-21 23:24 [Bridge] using bridges to connect vlans and untagged data Zoilo Gomez
2010-02-22  0:47 ` richardvoigt
2010-02-22  1:12   ` Zoilo Gomez
2010-02-22 16:04   ` james.card [this message]

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=C7A8154B.1201%james.card@nokia.com \
    --to=james.card@nokia.com \
    --cc=bridge@lists.linux-foundation.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