Networking issue ?

[conrad]

A few day a go i migrated al my servers to pve. All linux guest running on openvz and 2 windows guest on kvm.

On my samba guest the domain is not vissible anymore and i cannot join the domain. In nmbd.log is says that no local interfaces are found.
an tried a bridged interface, but cannot get that working, maybe it's a good idea to put a few steps in a howto howto configure a debian or centos vm for use with bridged networking.

when i go to eg. \\192.168.100.33 the samba share are visible.

i created a eth0 in the VM, but can only ping the VM itselfs. Do i have to do other steps to get
the networking set up ?

Thanks.

 

[tog]

I think you're looking at the right place and you'll be needing a bridged interface. The venet interface shows up in the guest container as a point-to-point interface with a netmask of 255.255.255.255 which will not work for nmbd which needs to broadcast.

Sorry I can't really help with getting a bridged interface running in a non-debian openvz guest container, haven't done it yet.

 

[conrad]

Thanks Tog,

At least i know now that i'm heading in the right direction

Every guest that wants to join a domain must have bridged networking i guess.

But i hope someone can tell me what setting/command i'm missing to get
my eth0 in Centos template working...

I do not use Centos often, but just this one is...

 

[conrad]

Also in a debian guest i created a eth0

on the host ifconfig -a displays a lo, vmbr0, veth110:0
The debian VM shows a lo, eth0, venet0 ??

the logfile on the host shows:
Configure veth devices: veth110.0
Adding interface veth110.0 to bridge vmbr0 on CT0 for CT110

Ik know i'm new to openvz, but i don't really understand the networking part.

 

[dietmar]

For Samba using a WINS Server should solve the problems.

When you use bridged networking, you configure your network like you are on real hardware.

- Dietmar

 

[conrad]

Ha Dietmar,

thanks for your reply, it wish it was so, but
i have a debian container with a normal eth0 en lo
en cannot ping anything besides itself.

Further a default installation of proxmox, and no strange settings
done on my part.
All venet vm's are working fine, just the bridged one for samba does not work.

 

[dietmar]

What is the content of /etc/network/interfaces?

 

[conrad]

my interfaces is just:

auto lo
iface lo inet loopback

auto eth0
iface eth0 inet static
address 192.168.100.32
netmask 255.255.255.0
gateway 192.168.100.200

 

[dietmar]

in /etc/vz/vznet.conf: change "/usr/bin/vznetaddbr" to "/usr/sbin/vznetaddbr"

and start the container again - does that solve the problem?

 

[conrad]

YES !!

that solved the problem, is that just a minor bug, strange that no-one else
mesioned it.

 

[dietmar]

I will release a bug fix soon.

- Dietmar

 

[conrad]

Thanks very much for your quick reply.... love proxmox!

 

[dietmar]

Ok, uploaded the fix. Please revert the changes I told you, and update the packages:

# apt-get update
# apt-get upgrade

- Dietmar

 

[conrad]

done upgrade

vznetaddbr is now in /usr/bin/ and /usr/sbin

ok?

 

[dietmar]

Thats OK (we only use '/usr/bin/vznetaddbr')

- Dietmar

 

[conrad]

All bridged VM have networking now.

Exept one that was migrated tot host2.
Also updated.

/etc/network/interfaces en routes looks the same,
but also no reply.

I have changed the venet to vmbr in the interface.
restarted the container.

Any idea what could be the reason for this.

 

[dietmar]

I have changed the venet to vmbr in the interface.
restarted the container.

And changed /etc/network/interfaces manually? Whats the content of /etc/network/interfaces?

 

[conrad]

exactly the same as on the VM on host 1

/etc/network/interfaces of 105 on host 2

auto lo
iface lo inet loopback

auto eth0
iface eth0 inet static
address 192.168.100.32
netmask 255.255.255.0
gateway 192.168.100.200

 

[conrad]

sorry 192.168.100.12 as ipadres (thats the only difference)

 

[dietmar]

and host2 is also a debian installation?

If so, what is the output of 'ifconfig' (on host and inside guest)?