You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It's a good question. I haven't tried it, but my guess is that they will both fight for the resources and allocation will fail.
I have tried it. It can work but needs lots of manual operations to change the GPU device driver from nvidia to vfio-pci or from vfio-pci to nvidia in the node.
Hi @caohuilong , I am also looking into this issue and is quite interested in allocating gpu to VM or pod in the same node. Could you share your setup about how to do the nvidia <-> vfio-pci driver switch? I wonder if a plugin in kubernetes could be made to make this process automatic.
Can i use kubevirt-gpu-device-plugin to allocate GPUs to Kubevirt VMs and use k8s-device-plugin to allocate GPUs to my container workloads in the same node?
The text was updated successfully, but these errors were encountered: