Add IP of another subnet

[dietmar]

Please test also without bonding device.

 

[charon]

same problem

perhaps it is a problem in the provider's switching/routing setup?

as i pointed out earlier, i can't see any arp replies on the host.

i'll ask the providers support if they see anything strange.
i am at my wits' end.

 

[charon]

Hosting support told me that i should activate a mac rewrite that tells the guest system to use the mac address of the physical interface.

first of all i will do a complete reinstallation.

 

[dietmar]

From what I know thats not possible with a bridge. Maybe you can just use routed network anyways (Do not connect eth0 to the bridge).

- Dietmar

 

[charon]

Sorry Dietmar i just do not get you right now.
i was trying different configurations the last few days.

Let me point it out:

OpenVZ Container venet -> works IN and OUT with route add and pointopoint
OpenVZ Container veth -> does not work
KVM vmbr -> does not work
KVM NAT -> works OUT (i do not know how to route INTO the KVM)

So is there a possible solution to clone the working OpenVZ configuration with KVM? In QEMU there is a -redirect option, but it does not work with proxmox.

Actually i dot not know how the different networking scenarios actually work. Perhaps it might be worth for the wiki if anyone can add an explanation?
Several people ask for this in the forums, but it also could be that i am too dumb to get it

I must admit, there is a problem with port security in the provider's switches (serverloft) and i do not know if he will change it.

Another idea would be to drop kvm virtualization in my case and get into creating the specific templates for openVZ. In my case i have to host several openSUSE machines and i can't convince the client of CentOS

charon

 

[dietmar]

I must admit, there is a problem with port security in the provider's switches (serverloft) and i do not know if he will change it.

Yes, the problem is that your provider block the packets if he sees packtes from different MAC addresses. So you simply cant connect vmbr0 directly to the network. You should use 'routed' network mode instead:

auto lo
iface lo inet loopback
 
iface eth0 inet static
       address  xxx.xxx.124.207
       netmask  255.255.255.192
       gateway  xxx.xxx.124.193
 
auto vmbr0
iface vmbr0 inet static
        address  xxx.xxx.127.129
        netmask  255.255.255.128
        bridge_ports none
        bridge_stp on
        bridge_fd 0

I assigned an additional IP address to the bridge (xxx.xxx.127.129), which you can use inside the VMs as gateway.

Now, packets arriving on vmbr0 are routed to eth0. So all packtet from your host use the eth0 mac address.

Got it?

- Dietmar

 

[dietmar]

Another idea would be to drop kvm virtualization in my case and get into creating the specific templates for openVZ. In my case i have to host several openSUSE machines and i can't convince the client of CentOS

Good Idea, but convince them to use Debian ;-)

 

[charon]

SOLVED!

Host:

qhost1:~# route -n
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
XXX.XXX.127.201   XXX.XXX.127.200   255.255.255.255 UGH   0      0        0 vmbr0
XXX.XXX.127.202   0.0.0.0         255.255.255.255 UH    0      0        0 venet0
XXX.XXX.124.192   0.0.0.0         255.255.255.192 U     0      0        0 eth0
XXX.XXX.127.128   0.0.0.0         255.255.255.128 U     0      0        0 vmbr0
0.0.0.0         XXX.XXX.124.193   0.0.0.0         UG    0      0        0 eth0
qhost1:~# cat /etc/network/interfaces
# network interface settings
auto lo
iface lo inet loopback

auto eth0
iface eth0 inet static
        address  XXX.XXX.124.207
        netmask  255.255.255.192
        gateway  XXX.XXX.124.193

iface eth1 inet manual

auto vmbr0
iface vmbr0 inet static
        address  XXX.XXX.127.200
        netmask  255.255.255.128
        bridge_ports none
        bridge_stp on
        bridge_fd 0

kvm

# The loopback network interface
auto lo
iface lo inet loopback
auto eth0
iface eth0 inet static
address XXX.XXX.127.201
netmask 255.255.255.128
gateway XXX.XXX.127.200

# route -n
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
XXX.XXX.127.128   0.0.0.0         255.255.255.128 U     0      0        0 eth0
0.0.0.0         XXX.XXX.127.200   0.0.0.0         UG    100    0        0 eth0