Low Availability VM Fallover

[jamesarbrown]

Is there a method for just 2 servers, low low availabilty.

Such if a VM is down on node1, it spins up the last storage replication and VM on node 2. Even if within 10 minutes will be good enough. The Vms only have kubernetes on there with all storage on an NFS server, so the images are pretty static.

Tried using HA, but with 2 nodes can confirm pulling the ethernet on one node infact crashed and rebooted the other...... didnt look too far into that as I know its not supported.

I guess I can use a Pi as Qdevice... but its just getting messy for no reason.

 

[shanreich]

Tried using HA, but with 2 nodes can confirm pulling the ethernet on one node infact crashed and rebooted the other...... didnt look too far into that as I know its not supported.

This is due to the node losing quorum and then fencing itself - you can read more about it in our docs [1]. This would be resolved by using a QDevice and is probably the most straightforward way of solving your problem.

[1] https://pve.proxmox.com/pve-docs/pve-admin-guide.html#ha_manager_fencing