Networking hung up in Proxmox Mail Gateway (as VM)

  • Thread starter Thread starter eugenyh
  • Start date Start date
E

eugenyh

Guest
So, have a Proxmox VE Cluster (1.8).

Setup KVM VM with Proxmox Mail Gateway (2.6, free license) as Email filter for our
corporate email server. Gateway and Email server behind firewall and have internal IP. Gateway published in internet via firewall.

KVM - because VE cluster have several VLAN interfaces,
and Proxmox MG must use one of them.

All working fine, but sometimes (each 1-2 days) my Proxmox MG stop to ping (and stop to receive and forward emails).

Im go to shell, do /etc/init.d/networking restart. And all start to working for next 1-2 days. Its very strange.

I have no such problems with my other VM.

For now I set network restart to cron every day at midnight.

But, I feel, its not a good :)

Any idea?
 
what kernel do you use inside the mail gateway? and post the VM config file (cat /etc/qemu-server/VMID.conf)
 
Kernel: 2.6.27-2-proxmox
Ver: 2.6/4540

Conf:
name: my-mailgw
ide2: vis01-vm-ISO-01:iso/proxmox-mailgateway_2.6.iso,media=cdrom
bootdisk: ide0
ostype: l26
ide0: vis01-vm-vdisks-01:109/vm-109-disk-1.raw
memory: 512
onboot: 1
sockets: 1
vlan1: rtl8139=CE:2D:18:2B:8D:A1
 
...
vlan1: rtl8139=CE:2D:18:2B:8D:A1

try e1000 instead of rtl8139, just change the line to:

Code:
vlan1: e1000=CE:2D:18:2B:8D:A1

stop and start again (a reboot of the PMG is not enough)