Good news come from Amos When I just want to report an issue about this nic
mode to both of you. Then I describe the issue for referring. I found the mac
addr generated
from code is same as the hwaddr in result of ifconfig in guest, but it
different from the hwaddr value in the ifcfg-eth0 file. So the framework report
error info of "Cannot find IP address for MAC address 9a:xx:xx:xx:xx:xx". In
last week, I report the pass of tap mode, because my image is setup manually.
Today, the case failed.
now, I will update code to use the new method of private bridge.
Thanks!
wuyu
On 2011-04-19,at 11:43:14, Lucas Meneghel Rodrigues wrote:
>发送日期：2011-04-19 11:43:14
>收件人：Amos Kong
>抄送：autotest
>主题：Re: [Autotest] [PATCH 1/3] KVM-test: Add script to setup privatebridge
>>
>On Tue, 2011-04-19 at 11:17 +0800, Amos Kong wrote:
>> Script is used to setup/remove of private bridge
>> it provides dhcp service to guests by dnsmasq.
>
>Ok, this is a great start! I'd rather have this as python code as part
>of the framework, so I'll play with it a little and we can get this
>going.
>
>Jason, would you please re-send that patchset that removes the need for
>the qemu-ifup script? We can make everything work together and deliver a
>good experience for users and completely getting rid of user mode for
>ever and ever :)
>
>> Signed-off-by: Amos Kong <ak...@redhat.com>
>> ---
>> client/tests/kvm/scripts/set_private_br.sh | 53
>> ++++++++++++++++++++++++++++
>> 1 files changed, 53 insertions(+), 0 deletions(-)
>> create mode 100755 client/tests/kvm/scripts/set_private_br.sh
>>
>> diff --git a/client/tests/kvm/scripts/set_private_br.sh
>> b/client/tests/kvm/scripts/set_private_br.sh
>> new file mode 100755
>> index 0000000..4c4cff0
>> --- /dev/null
>> +++ b/client/tests/kvm/scripts/set_private_br.sh
>> @@ -0,0 +1,53 @@
>> +#!/bin/bash
>> +#
>> +# Author: Amos Kong <kongjian...@gmail.com>
>> +# Date: Apr 14, 2011
>> +# Description: this script is used to create/delete a private bridge,
>> +# launch a dhcp server on the bridge by dnsmasq.
>> +#
>> +# @ ./set_private_br.sh $bridge_name $subnet_prefix
>> +# @ ./set_private_br.sh vbr0 192.168.33
>> +
>> +brname='vbr0'
>> +subnet='192.168.33'
>> +
>> +add_br()
>> +{
>> + echo "add new private bridge: $brname"
>> + /usr/sbin/brctl show |egrep "\b$brname\b" && exit
>> + /usr/sbin/brctl addbr $brname
>> + echo 1 > /proc/sys/net/ipv4/ip_forward
>> + /usr/sbin/brctl stp $brname on
>> + /usr/sbin/brctl setfd $brname 0
>> + ifconfig $brname $subnet.1 up
>> + # Add forward rule, then guest can access public network
>> + iptables -t nat -A POSTROUTING -s $subnet.254/24 ! -d $subnet.254/24 -j
>> MASQUERADE
>> + /etc/init.d/dnsmasq stop
>> + dnsmasq --strict-order --bind-interfaces --listen-address $subnet.1
>> --dhcp-range $subnet.1,$subnet.254 --pid-file=/tmp/dnsmasq.pid
>> +}
>> +
>> +del_br()
>> +{
>> + echo "cleanup bridge setup"
>> + kill -9 `cat /tmp/dnsmasq.pid`
>> + ifconfig $brname down
>> + /usr/sbin/brctl delbr $brname
>> + iptables -t nat -D POSTROUTING -s $subnet.254/24 ! -d $subnet.254/24 -j
>> MASQUERADE
>> +}
>> +
>> +
>> +if [ $# = 0 ]; then
>> + echo "./set_private_br.sh add|del [brname] [subnet]"
>> + exit
>> +fi
>> +if [ $# > 1 ]; then
>> + brname="$2"
>> +fi
>> +if [ $# > 2 ]; then
>> + subnet="$3"
>> +fi
>> +if [ $1 = 'add' ]; then
>> + add_br
>> +elif [ $1 = 'del' ]; then
>> + del_br 2>/dev/null
>> +fi
>>
>
>
>_______________________________________________
>Autotest mailing list
>Autotest@test.kernel.org
>http://test.kernel.org/cgi-bin/mailman/listinfo/autotest
>.
>
------------------
wuyu8242
2011-04-19
_______________________________________________
Autotest mailing list
Autotest@test.kernel.org
http://test.kernel.org/cgi-bin/mailman/listinfo/autotest
