Proxmox with NFS Gateway is breaking incoming migrations

[marcellinus77]

Following the tutorial: https://linbit.com/blog/highly-available-nfs-for-proxmox-with-linstor-gateway/ for setting up basic NFS share on Proxmox.

Having the virtual IP on node in the cluster is breaking incoming mirgation on this node:

could not get migration ip: multiple, different, IP address configured for network '10.1.2.11/24' TASK ERROR: command '/usr/bin/ssh -e none -o 'BatchMode=yes' -o 'HostKeyAlias=node-2' -o 'UserKnownHostsFile=/etc/pve/nodes/node-2/ssh_known_hosts' -o 'GlobalKnownHostsFile=none' root@10.1.2.12 pvecm mtunnel -migration_network 10.1.2.11/24 -get_migration_ip' failed: exit code 255

while cluster sonsits of nodes:

node-1 10.1.2.11/24
node-2 10.1.2.12/24
node-3 10.1.2.13/24
node-4 10.1.2.14/24

and VIP is set to 10.1.2.10/24 (currently running on node-2)

vmbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 qdisc noqueue state UP group default qlen 1000
    link/ether XX:XX:XX:XX:XX:XX brd ff:ff:ff:ff:ff:ff
    inet 10.1.2.12/24 scope global vmbr0
       valid_lft forever preferred_lft forever
    inet 10.1.2.10/24 scope global secondary vmbr0
       valid_lft forever preferred_lft forever
    inet6 fe80::aaaa:bbbb:cccc:dddd/64 scope link 
       valid_lft forever preferred_lft forever

Kr M.

[chrboe]

I don't know too much about Proxmox, so I'm only guessing here, but based on this forum post, it seems like Proxmox dislikes a node having two IP addresses in the same subnet.

So you could try setting up your "NFS subnet" to be different from your "migration subnet", for example 10.2.2.0/24.