OpenVZ on PVE 3.4-6 fails on multiple subnet

[apmuthu]

PVE 3.4-6

LAN: 192.168.1.0/24 => eth0 => vmbr0 (with gateway)

WAN: 1.2.3.4/28 => eth1 => vmbr1

LAN2: 192.168.2.0/24 => No physical NIC (none) => vmbr2

No special Post ups - all bridges created and configured within PVE 3.4-6

Create OpenVZ VM 101 on LAN2 (192.168.2.20/24) using venet
Cannot start it.,
Error:

Starting container ...
Container is mounted
Adding IP address(es): 192.168.2.20
arpsend: can't get iface 'vmbr1' address : Cannot assign requested address
vps-net_add ERROR: arpsend -c 1 -w 1 -D -e 192.168.2.20 vmbr1 FAILED
Container start failed (try to check kernel messages, e.g. "dmesg | tail")
Killing container ...
Container was stopped
Container is unmounted
TASK ERROR: command 'vzctl start 101' failed: exit code 1

Instead of connecting to vmbr2, it is attempting to connect to vmbr1.

 

[apmuthu]

SOLVED:

The file /etc/vz/vz.conf has a safe value by default as:

## Fail if there is another machine in the network with the same IP
ERROR_ON_ARPFAIL="no"

I had set that parameter to "yes" and that caused all the problem.
Reverted back to the default and all is well.