Vm creation error with gpu

[TheKunalSen]

while i am creating a vm using the gpu i am getting this error
i am using rtx3050 for this

2026-03-20 07:41:46,278 WARN  [resource.wrapper.LibvirtStartCommandWrapper] (AgentRequest-Handler-3:[]) (logid:62d6f5fb) LibvirtException org.libvirt.LibvirtException: unsupported configuration: host doesn't support VFIO PCI passthrough
        at org.libvirt.ErrorHandler.processError(Unknown Source)
        at org.libvirt.ErrorHandler.processError(Unknown Source)
        at org.libvirt.Connect.domainCreateXML(Unknown Source)
        at com.cloud.hypervisor.kvm.resource.LibvirtComputingResource.startVM(LibvirtComputingResource.java:2241)
        at com.cloud.hypervisor.kvm.resource.LibvirtComputingResource.startVM(LibvirtComputingResource.java:2210)
        at com.cloud.hypervisor.kvm.resource.wrapper.LibvirtStartCommandWrapper.execute(LibvirtStartCommandWrapper.java:91)
        at com.cloud.hypervisor.kvm.resource.wrapper.LibvirtStartCommandWrapper.execute(LibvirtStartCommandWrapper.java:52)
        at com.cloud.hypervisor.kvm.resource.wrapper.LibvirtRequestWrapper.execute(LibvirtRequestWrapper.java:78)
        at com.cloud.hypervisor.kvm.resource.LibvirtComputingResource.executeRequest(LibvirtComputingResource.java:2280)
        at com.cloud.agent.Agent.processRequest(Agent.java:813)
        at com.cloud.agent.Agent$AgentRequestHandler.doTask(Agent.java:1295)
        at com.cloud.utils.nio.Task.call(Task.java:83)
        at com.cloud.utils.nio.Task.call(Task.java:29)
        at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:264)
        at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1128)
        at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:628)
        at java.base/java.lang.Thread.run(Thread.java:829)

user@kvm-host-gpu-1:~$ lspci -nnk | grep -i nvidia -A3
01:00.0 VGA compatible controller [0300]: NVIDIA Corporation Device [10de:2582] (rev a1)
        Subsystem: ZOTAC International (MCO) Ltd. Device [19da:2641]
        Kernel modules: nvidiafb, nouveau
01:00.1 Audio device [0403]: NVIDIA Corporation Device [10de:2291] (rev a1)
        Subsystem: ZOTAC International (MCO) Ltd. Device [19da:2641]
        Kernel driver in use: snd_hda_intel
        Kernel modules: snd_hda_intel

[vishesh92]

@TheKunalSen Looks like the host is not configured correctly. Check the cwiki doc for more details on how to configure the host.

Foe EL, you can check point 1 & 2 under 16.1 in redhat's documentation on how to configure the host.

let me know if this resolves your issue

[TheKunalSen]

yes it worked but i have a question right now the gpu is getting allocated to only one vm, is there any way where i can use vgpu like we do using the Cpu overprovisioning factor.

[vishesh92]

This depends on the GPU card. If it supports partitioning, it should work. Only NVIDIA vGPU is supported at this time.

For NVIDIA RTX 3050, AFAIK it is a consumer GPU so it doesn't support vGPUs out of the box. There are some community based projects which allow usage of vgpu on consumer GPUs but it doesn't support all GPUs and is not something which is provided officially by NVIDIA.

[TheKunalSen]

okay thank you