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

[Xen-users] Problem of slowness network when Nash is killed

To: xen-users@xxxxxxxxxxxxxxxxxxx
Subject: [Xen-users] Problem of slowness network when Nash is killed
From: Michel Desconnets <michel.desconnets@xxxxxxxxx>
Date: Tue, 13 Nov 2007 20:46:59 +0100
Delivery-date: Tue, 13 Nov 2007 11:48:19 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
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>
Sender: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: Mozilla-Thunderbird 2.0.0.6 (X11/20071009)
(sorry for my english)
I have a strange problem :

I use for my dom0 the kernel of dom0
On my DomU (all is redhat 4.3), Process nash-hotplug take 99% CPU.
I have already see in mail archives that i can kill this process.
Indeed, is a good soluce, but It have big consequence : Network become very slow !

In an other machine on network, i use a ping flood and i look performance.
When nash is running, i observe 200-300 ko/s.
If i kill nash (or just STOP it with a kill -STOP), i have 40-50 ko/s maximum (if i restart nash with kill -CONT, i have again 300ko/s !)

I try to make a new initrd in domU : i don't have nash any more, but network is slow every time.

I notice that if i use vcpus > 1 in domU, network is slow every time.
with vpcus=1 and nash running : all is ok (except cpu load)

I have 2 others servers, exactly the same, but without several bridge (just one), upgrade ok for centos 5, and network stay quick when i kill nash: it is a coincidence, or a bug is already fixed ?

DomU : redhat 4.3, 64bit
ex :
kernel = '/boot/vmlinuz-2.6.18-8.1.4.el5xen'
ramdisk = '/boot/initrd-2.6.18-8.1.4.el5xen-domU.img'
memory  = '750'
root ='LABEL=xenpiraatsda1'
disk    = ['phy:/dev/rootvg/xen_piraat_sda1,sda1,w']
cpus = "1-3"
vcpus   = 1
#  Hostname
name    = 'piraat'
#  Networking
dhcp = 'dhcp'
vif  = [  'mac=00:16:3E:2F:18:F7, bridge=xenbr0' ]
#  Behaviour

on_reboot   = 'restart'
on_crash    = 'restart'



Dom0 :
CentOS release 5 (Final) 64bit, 16Go de RAM
Linux helicon 2.6.18-8.1.4.el5xen #1 SMP Thu May 17 03:43:13 EDT 2007 x86_64 x86_64 x86_64 GNU/Linux
xen-3.0.3-25.0.4.el5


2 bridges for network (1 per vlan) :
# !/bin/sh
# network-xen-multi-bridge
# Exit if anything goes wrong
set -e
# First arg is operation.
OP=$1
shift
script=/etc/xen/scripts/network-bridge.xen
case ${OP} in
start)
$script start vifnum=1 bridge=xenbr1 netdev=eth1
$script start vifnum=0 bridge=xenbr0 netdev=eth0
;;
stop)
$script stop vifnum=1 bridge=xenbr1 netdev=eth1
$script stop vifnum=0 bridge=xenbr0 netdev=eth0
;;
status)
$script status vifnum=1 bridge=xenbr1 netdev=eth1
$script status vifnum=0 bridge=xenbr0 netdev=eth0
;;
*)
echo 'Unknown command: ' ${OP}
echo 'Valid commands are: start, stop, status'
exit 1
esac


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