[CentOS-virt] Centos 6.2 / KVM troubles with network

Sat Jan 14 14:41:32 UTC 2012
Radek Bursztynowski <radek at bursztynowski.waw.pl>

Dnia 2012-01-14, sob o godzinie 15:08 +0100, Dennis Jacobfeuerborn
pisze:

> On 01/14/2012 02:59 PM, Radek Bursztynowski wrote:
> > Hi,
> >
> > I have installed CentOS 6.2 x86_64 and KVM/Qemu and I have some troubles
> > with network.
> >
> > The first case:
> > -----------------
> >
> > The configuration is:
> > - CentOS 6.2 (hardware machine) has one Ethernet interface with static IP
> > address (eth0).
> > - Virtual machine has bridge network interface with eth0 interface on
> > CentOS 6.2. This bridge is created using wizard integrated with virt-manager.
> > - Bridge interface on virtual machine receives IP addres from DHCP server.
> > IP address is received well.
> > - All firewalls and SELinux are closed.
> >
> > Virtual machine has no any connection with CentOS 6.2 hardware machine, and
> > CentOS 6.2 has no any connection with virtual machine. But every others
> > machines have all connection with both virtual and CentOS 6.2 machine.
> >
> > The question: Why?
> >
> > The second case:
> > ----------------------
> >
> > The configuration is:
> > - On CentOS 6.2 (hardware machine) I created bridhe (br0) with eth0
> > hardware interface. Two situations:
> > + I force two different IP addresses for eth0 and br0 interfaces - there no
> > any connection with this server. So I don't test this case.
> > + I forced one IP addres for both eth0 and br0 interfaces. ifconfig shows
> > that eth0 has no IP address, br0 has IP address. So, I test this case.
> > - I use Red Hat documentation:
> > http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/5/html-single/Virtualization/index.html#sect-Virtualization-Network_Configuration-Bridged_networking_with_libvirt)
> > - On virtual machine I created bridge connection with br0 interface on
> > CentOS 6.2.
> > - All machines have all connections.
> > - All firewalls and SELinux are closed.
> >
> > NFS server on CentOS 6.2 (hardware machine) doesn't serve exported file system.
> >
> > The questions:
> > 1. Is it proper that eth0 has no IP addres and bridge (br0) interface has?
> > 2. What about my NFS? NFS server worked well before starting virtualization.
> >
> > Who can me help with my troubles?
> 
> You say you create bridge interfaces in your virtual machines. Why? Usually 
> you don't do that you create a bridge only on the host.
> 
> You also say you "forced" one IP address for both eth0 and br0. There is no 
> need to force anything. eth0 shouldn't have an IP but br0 should have this 
> IP instead.
> 
> Regards,
>    Dennis
> _______________________________________________

Dennis,

Thanks for interesting. Few words explanation. 
I have one network segment, and I try to joint virtual machines to this
segment. So my hardvare machine has 192.168.0.1 address, gateway is
192.168.0.254, netmask = 255.255.255.0.

Regarding bridge on my virtual machine. When I started virtualization I
had NAT only in another network segment. So, I look for solution adding
virtual machines to my base network segment. Previously I used VMware
Server, and it was proper solution.

Regarding my force one IP addres for both interface. I used this
solution because eth0 has no any IP addres and I couldn't find CentOS
6.2 server in the network. Using exactly Red Hat configuration I have:

# more ifcfg-eth0
DEVICE=eth0
TYPE=Ethernet
BOOTPROTO=none
IPADDR=192.168.0.1
NETMASK=255.255.255.0
PREFIX=24
GATEWAY=192.168.0.254
DNS1=192.168.0.199
DNS2=194.204.159.1
DNS3=194.204.152.34
DEFROUTE=yes
IPV4_FAILURE_FATAL=yes
IPV6INIT=no
NAME=eth0
ONBOOT=yes
BRIDGE=br0
MTU=9000
DOMAIN=bursztynowski.waw.pl
UUID=5fb06bd0-0bb0-7ffb-45f1-d6edd65f3e03

# more ifcfg-br0
DEVICE=br0
TYPE=Bridge
BOOTPROTO=dhcp
#IPADDR=192.168.0.1
#NETMASK=255.255.255.0
#PREFIX=24
#GATEWAY=192.168.0.254
#DNS1=192.168.0.199
#DNS2=194.204.159.1
#DNS3=194.204.152.34
ONBOOT=yes
DELAY=0

(previous configuration br0 was: 
BOOTPRO=static
and with no '#' characters. Then all worked excluding NFS.


# ifconfig                            
br0       Link encap:Ethernet  HWaddr 00:50:8D:94:2F:B7            
          inet addr:192.168.0.92  Bcast:192.168.0.255
Mask:255.255.255.0
          inet6 addr: fe80::250:8dff:fe94:2fb7/64
Scope:Link             
          UP BROADCAST RUNNING MULTICAST  MTU:9000
Metric:1             
          RX packets:16 errors:0 dropped:0 overruns:0
frame:0            
          TX packets:26 errors:0 dropped:0 overruns:0
carrier:0          
          collisions:0
txqueuelen:0                                      
          RX bytes:3034 (2.9 KiB)  TX bytes:4313 (4.2
KiB)               

eth0      Link encap:Ethernet  HWaddr 00:50:8D:94:2F:B7  
          inet6 addr: fe80::250:8dff:fe94:2fb7/64 Scope:Link
          UP BROADCAST RUNNING PROMISC MULTICAST  MTU:9000  Metric:1
          RX packets:57255 errors:0 dropped:0 overruns:0 frame:0
          TX packets:60263 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:21481457 (20.4 MiB)  TX bytes:15768417 (15.0 MiB)
          Interrupt:23 Base address:0x8000

lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:30723 errors:0 dropped:0 overruns:0 frame:0
          TX packets:30723 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:92725098 (88.4 MiB)  TX bytes:92725098 (88.4 MiB)

Should I replace static and dhcp addresses between eth0 and br0
interface?

Regards,
Radek






-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.centos.org/pipermail/centos-virt/attachments/20120114/bd769edd/attachment-0006.html>