Transfer to second node - networking fails?

[anthony]

All, I had a node fail in our HA Cluster. When the VMs transferred over to another node the networking failed. after manually rebooting each vm the networking was up and operational. I am not using any advanced proxmox features like the firewall or anything. The bridges have the same ID and and all that. they were a mix of paravirtualized and e1000.

when logged into the vm the interface appears as up, however when trying to ping anything, I get 0 traffic. What could cause this? is there troubleshooting steps one may suggest?

Thanks!!

 

[oguz]

hi,

if you had a different network setup on the other node you transferred to, it is likely that your VM failed to assign an IP or it may be a routing problem.

check the network configuration inside the VM once again to make sure everything is correct. (IP, gateway, netmask, dns etc.)

is there troubleshooting steps one may suggest?

you can use tcpdump to watch the network packets and debug.

 

[spirit]

seem to be an arp problem.

with virtio-net interface, vm should send gratuitous arp after the live migration.
for e1000, I'm not sure about the behaviour.

Do you use bonding for your proxmox hosts ? if yes, which mode ?