Unable to install 3D driver on VM on RedHat 7.9 OS version

The 470 driver is now active, but

NVRM: installed in this system is not supported by the
NVRM: NVIDIA 470.74 driver release.
NVRM: Please see ‘Appendix A - Supported NVIDIA GPU Products’
NVRM: in this release’s README, available on the operating system
NVRM: specific graphics driver download page at www.nvidia.com.

This doesn’t mean that your gpu is unsupported but you’re running an unsupported VM config. (unsupported hypervisor, consumer card, vm boot) Since VMWare + Tesla is supported, your VM and/or host bios is incorrectly set up, I suspect. Looking at the boot messages, you’re running a bios boot with 32bit resources, please see this on how to correctly set up a VM for passthrough:
https://blogs.vmware.com/apps/2018/09/using-gpus-with-virtual-machines-on-vsphere-part-2-vmdirectpath-i-o.html

verified the vm config and we are using shared pci. but am able to installing on rhel 7.6 but 7.9 am getting the same error

nvidia-smi
Tue Nov 9 08:34:23 2021
±----------------------------------------------------------------------------+
| NVIDIA-SMI 390.115 Driver Version: 390.115 |
|-------------------------------±---------------------±---------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GRID P6-2B On | 00000000:02:01.0 Off | N/A |
| N/A N/A P8 N/A / N/A | 144MiB / 2048MiB | 0% Default |
±------------------------------±---------------------±---------------------+

±----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| No running processes found |
±----------------------------------------------------------------------------+
[root@awv344802 nxf43580]# cat /etc/redhat-release
Red Hat Enterprise Linux Server release 7.6 (Maipo)

I only now noticed that you’re running a vGPU setup (GRID P6-2B), please try using the grid driver.

could you please share the grid driver version which support rhel 7u9

Those should be available at the licensing portal:
https://www.nvidia.com/en-us/drivers/vgpu-software-driver/

ok. but how it’s works for rhel 7u6 because i used the same grid version for rhel 7u9 NVIDIA-Linux-x86_64-390.115-grid.run

In your initial post you had compiler errors, the 390.115 likely being too old to compile against the centos 7.9 kernel. Isn’t there a 460-grid driver available from the portal?

am unable to login the portal. we have our organization account let me try with that.
460-grid driver will resolve this issue ?

It should do the trick
v390.115 released mid-2019
centos 7.9 late-2020
v460 late-2020, after centos 7.9


this one ?

It also depends on the vgpu manager version and hypervisor you’re running. I don’t know which version you have but you said you’re running on vmware. Please check for vmware downloads and which vgpu manager version you’re running.

vsphere 6.7 using

before we have vpshere 6.0 and recently we migrated to 6.7

VIDIA-vGPU-VMware_ESXi_6.0_Host_Driver 390.113-1OEM.600.0.0.2494585 NVIDIA VMwareAccepted 2019-05-31

so do i need to reconfigure the esxi with latest ?

According to
https://docs.nvidia.com/grid/get-grid-version.html
You seem to be running vgpu 6.4 which is really old so you need to upgrade that in order to have centos 7.9 support.
The grid drivers only run on manager versions equal or higher.
https://docs.nvidia.com/grid/latest/pdf/grid-vgpu-release-notes-red-hat-el-kvm.pdf

You might check if the normal 390.144 graphics driver works on your vpu setup, but I don’t expect that.
https://http.download.nvidia.com/XFree86/Linux-x86_64/390.144/

yes it woks only for 7u6 but not 7u9

Out of curiosity, what errors do the 390.144 give on 7.9 (dmesg output)?

support.txt (122.7 KB)

nvidia-smi vgpu
Not supported devices in vGPU mode

why it’s showing this error

Please check if this applies:
https://forums.developer.nvidia.com/t/solved-m10-with-esxi-6-5-vgpu-device-not-supported/162576/7?u=generix

one more observation.

gpuvm
Xserver unix:0, PCI ID 0:92:0:0, vGPU: Not set, GPU maximum memory 16738636KB
vgpu not set