We have migrated a few Windows 2003 server VMs running on Nx / ESX to a new AHV cluster but in one of them after 3-4 hours the principal guest VM NIC stops responding and the VM has to be rebooted. We have tried deleting and adding the NIC manually with the same result so we had to roll back and boot that specific VM on the old Nutanix / ESX cluster.
We plan to upgrade Fedora drivers to latest instead of stable and maybe change some network registry settings according to W2k3 best practices.
Has anyone experienced this before? Any advice will be welcome.