WARNING - OLD ARCHIVES

This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-users

[Re: [Xen-users] Slow domU network (even within domU)]

To: xen-users <xen-users@xxxxxxxxxxxxxxxxxxx>
Subject: [Re: [Xen-users] Slow domU network (even within domU)]
From: Artur Schiefer <aschiefer@xxxxxx>
Date: Thu, 09 Mar 2006 17:00:17 +0100
Delivery-date: Thu, 09 Mar 2006 16:01:22 +0000
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-users-request@lists.xensource.com?subject=help>
List-id: Xen user discussion <xen-users.lists.xensource.com>
List-post: <mailto:xen-users@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=unsubscribe>
Organization: VUB Printmedia GmbH
Reply-to: aschiefer@xxxxxx
Sender: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
hello,

thank you but that didn't help neither. another strange observation i
made is when do i an iperf within the domu i get the following results:

iperf to localhost      2.8 gbit
iperf to domu-real-ip   200 mbit

(see results at the end).

ciao,
artur

Server listening on TCP port 5001
TCP window size: 85.3 KByte (default)
------------------------------------------------------------
VM_OTRS:~# iperf -c 192.168.70.50
------------------------------------------------------------
Client connecting to 192.168.70.50, TCP port 5001
TCP window size: 64.0 KByte (default)
------------------------------------------------------------
[  3] local 192.168.70.50 port 41309 connected with 192.168.70.50 port
5001
[  4] local 192.168.70.50 port 5001 connected with 192.168.70.50 port
41309
[  3]  0.0-10.0 sec    239 MBytes    200 Mbits/sec
[  4]  0.0-10.0 sec    239 MBytes    200 Mbits/sec
VM_OTRS:~# iperf -c 127.0.0.1
------------------------------------------------------------
Client connecting to 127.0.0.1, TCP port 5001
TCP window size: 64.0 KByte (default)
------------------------------------------------------------
[  3] local 127.0.0.1 port 56978 connected with 127.0.0.1 port 5001
[  4] local 127.0.0.1 port 5001 connected with 127.0.0.1 port 56978
[  3]  0.0-10.0 sec  3.25 GBytes  2.79 Gbits/sec
[  4]  0.0-10.0 sec  3.25 GBytes  2.79 Gbits/sec

Am Donnerstag, den 09.03.2006, 09:14 -0500 schrieb Himanshu Raj:
> You will have to tune your TCP parameters like window sizes etc. I
don't have a
> ready ref on this - but try the following settings in your domU.
> 
> Following sysctl parameters must be tuned in order to get gigE bw from
domUs.
> Either set via sysctl -w or put in /etc/sysctl.conf.
> 
> # increase TCP maximum buffer size
> net.core.rmem_max=16777216
> net.core.wmem_max=16777216
> 
> # increase Linux autotuning TCP buffer limits
> # min, default, and maximum number of bytes to use
> net.ipv4.tcp_rmem="4096 87380 16777216"
> net.ipv4.tcp_wmem="4096 65536 16777216"
> 
> -Himanshu
> 
> On Wed, Mar 08, 2006 at 02:27:29PM +0100, Artur Schiefer wrote:
> > hello philipp,
> > 
> > this is my xentop output when i am runnig the test. do you achieve
> > higher rates (should only be limited by memory throughtput).
> > 
> > ciao,
> > artur
> > 
> > xentop - 14:21:58   Xen 3.0.1
> > 2 domains: 1 running, 0 blocked, 0 paused, 0 crashed, 0 dying, 0
> > shutdown
> > Mem: 1048104k total, 1039148k used, 8956k free    CPUs: 4 @ 2799MHz
> >       NAME  STATE   CPU(sec) CPU(%)     MEM(k) MEM(%)  MAXMEM(k)
> > MAXMEM(%) VCPUS NETS NETTX(k) NETRX(k) SSID
> >   Domain-0 -----r       1170  109.3     885948   84.5   no limit
> > n/a     4    8        0        0    0
> >  vm_apache ------        690   72.4     131068   12.5     139264
> > 13.3     2    1  1196352    31262    0
> > Am Mittwoch, den 08.03.2006, 12:52 +0100 schrieb Philipp Jggi:
> > > 
> > > Did you recoded the xentop output during your test? How much
mem-max
> > > do you have for dom0 and your domU? 
> > > 
> > > bye Philipp 
> > > 
> > > 
> > > 
> > > 
> > > Artur Schiefer <aschiefer@xxxxxx> 
> > > Sent by:
> > > xen-users-bounces@xxxxxxxxxxxxxxxxxxx 
> > > 
> > > 03/08/2006 12:47 PM 
> > >          Please respond to
> > >          aschiefer@xxxxxx
> > > 
> > > 
> > > 
> > > 
> > >                To
> > > xen-users
> > > <xen-users@xxxxxxxxxxxxxxxxxxx> 
> > >                cc
> > > 
> > >           Subject
> > > [Xen-users] Slow
> > > domU network
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > > hello,
> > > 
> > > when i test network performance between dom0 <-> domU (xen
3.0.1tha3
> > > debian patches)with iperf i get only about 450 Mbits of throughput
and
> > > lots of dropped packets on the vif-interface (bridged). In
opposite to
> > > that, when i run the same test beetween two dom0 with
bonded/teamed
> > > gigabit-nics i am able to achieve 1.6 Gbits of throughput.
> > > has someone made this obsevation as well (what is your
throughput)?
> > > any solutions ?
> > > 
> > > cheers,
> > > artur
> > > 
> > > 
> > > 
> > > 
> > > _______________________________________________
> > > Xen-users mailing list
> > > Xen-users@xxxxxxxxxxxxxxxxxxx
> > > http://lists.xensource.com/xen-users
> > > 
> > 
> > 
> > _______________________________________________
> > Xen-users mailing list
> > Xen-users@xxxxxxxxxxxxxxxxxxx
> > http://lists.xensource.com/xen-users
> 

--- Begin Message ---
To: Himanshu Raj <rhim@xxxxxxxxxxxxx>
Subject: Re: [Xen-users] Slow domU network (even within domU)
From: Artur Schiefer <aschiefer@xxxxxx>
Date: Thu, 09 Mar 2006 16:54:13 +0100
In-reply-to: <20060309141446.GE2622@xxxxxxxxxxxxx>
Organization: VUB Printmedia GmbH
References: <OFA03D19D3.C8851746-ONC125712B.00411B0C-C125712B.00416B65@xxxxxxx> <1141824449.19645.17.camel@stella> <20060309141446.GE2622@xxxxxxxxxxxxx>
Reply-to: aschiefer@xxxxxx
hello,

thank you but that didn't help neither. another strange observation i
made is when do i an iperf within the domu i get the following results:

iperf to localhost      2.8 gbit
iperf to domu-real-ip   200 mbit

(see results at the end).

ciao,
artur

Server listening on TCP port 5001
TCP window size: 85.3 KByte (default)
------------------------------------------------------------
VM_OTRS:~# iperf -c 192.168.70.50
------------------------------------------------------------
Client connecting to 192.168.70.50, TCP port 5001
TCP window size: 64.0 KByte (default)
------------------------------------------------------------
[  3] local 192.168.70.50 port 41309 connected with 192.168.70.50 port
5001
[  4] local 192.168.70.50 port 5001 connected with 192.168.70.50 port
41309
[  3]  0.0-10.0 sec    239 MBytes    200 Mbits/sec
[  4]  0.0-10.0 sec    239 MBytes    200 Mbits/sec
VM_OTRS:~# iperf -c 127.0.0.1
------------------------------------------------------------
Client connecting to 127.0.0.1, TCP port 5001
TCP window size: 64.0 KByte (default)
------------------------------------------------------------
[  3] local 127.0.0.1 port 56978 connected with 127.0.0.1 port 5001
[  4] local 127.0.0.1 port 5001 connected with 127.0.0.1 port 56978
[  3]  0.0-10.0 sec  3.25 GBytes  2.79 Gbits/sec
[  4]  0.0-10.0 sec  3.25 GBytes  2.79 Gbits/sec

Am Donnerstag, den 09.03.2006, 09:14 -0500 schrieb Himanshu Raj:
> You will have to tune your TCP parameters like window sizes etc. I don't have 
> a
> ready ref on this - but try the following settings in your domU.
> 
> Following sysctl parameters must be tuned in order to get gigE bw from domUs.
> Either set via sysctl -w or put in /etc/sysctl.conf.
> 
> # increase TCP maximum buffer size
> net.core.rmem_max=16777216
> net.core.wmem_max=16777216
> 
> # increase Linux autotuning TCP buffer limits
> # min, default, and maximum number of bytes to use
> net.ipv4.tcp_rmem="4096 87380 16777216"
> net.ipv4.tcp_wmem="4096 65536 16777216"
> 
> -Himanshu
> 
> On Wed, Mar 08, 2006 at 02:27:29PM +0100, Artur Schiefer wrote:
> > hello philipp,
> > 
> > this is my xentop output when i am runnig the test. do you achieve
> > higher rates (should only be limited by memory throughtput).
> > 
> > ciao,
> > artur
> > 
> > xentop - 14:21:58   Xen 3.0.1
> > 2 domains: 1 running, 0 blocked, 0 paused, 0 crashed, 0 dying, 0
> > shutdown
> > Mem: 1048104k total, 1039148k used, 8956k free    CPUs: 4 @ 2799MHz
> >       NAME  STATE   CPU(sec) CPU(%)     MEM(k) MEM(%)  MAXMEM(k)
> > MAXMEM(%) VCPUS NETS NETTX(k) NETRX(k) SSID
> >   Domain-0 -----r       1170  109.3     885948   84.5   no limit
> > n/a     4    8        0        0    0
> >  vm_apache ------        690   72.4     131068   12.5     139264
> > 13.3     2    1  1196352    31262    0
> > Am Mittwoch, den 08.03.2006, 12:52 +0100 schrieb Philipp Jggi:
> > > 
> > > Did you recoded the xentop output during your test? How much mem-max
> > > do you have for dom0 and your domU? 
> > > 
> > > bye Philipp 
> > > 
> > > 
> > > 
> > > 
> > > Artur Schiefer <aschiefer@xxxxxx> 
> > > Sent by:
> > > xen-users-bounces@xxxxxxxxxxxxxxxxxxx 
> > > 
> > > 03/08/2006 12:47 PM 
> > >          Please respond to
> > >          aschiefer@xxxxxx
> > > 
> > > 
> > > 
> > > 
> > >                To
> > > xen-users
> > > <xen-users@xxxxxxxxxxxxxxxxxxx> 
> > >                cc
> > > 
> > >           Subject
> > > [Xen-users] Slow
> > > domU network
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > > hello,
> > > 
> > > when i test network performance between dom0 <-> domU (xen 3.0.1tha3
> > > debian patches)with iperf i get only about 450 Mbits of throughput and
> > > lots of dropped packets on the vif-interface (bridged). In opposite to
> > > that, when i run the same test beetween two dom0 with bonded/teamed
> > > gigabit-nics i am able to achieve 1.6 Gbits of throughput.
> > > has someone made this obsevation as well (what is your throughput)?
> > > any solutions ?
> > > 
> > > cheers,
> > > artur
> > > 
> > > 
> > > 
> > > 
> > > _______________________________________________
> > > Xen-users mailing list
> > > Xen-users@xxxxxxxxxxxxxxxxxxx
> > > http://lists.xensource.com/xen-users
> > > 
> > 
> > 
> > _______________________________________________
> > Xen-users mailing list
> > Xen-users@xxxxxxxxxxxxxxxxxxx
> > http://lists.xensource.com/xen-users
> 

--- End Message ---
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
<Prev in Thread] Current Thread [Next in Thread>