RHEL 7 Maipo server doesn't load nvidia driver 450.80.02 (4 X Tesla V100) after reboot

I’ve updated new NVidia driver (450.80.02) on RHEL 7 Maipo server with 4 Tesla’s V100. However, after reboot driver was not loaded. And nvidia-smi can’t connect to driver.
You may find logs in the attachment.
nvidia-bug-report.log.gz (66.7 KB)
Could you help me to figure out how to solve this issue?

The log is from a system with a T4 running ubuntu. Wrong file attached?

generix, sorry. I’ve updated the attachment. Now the file is correct.

The kernel modules were compiled and installed fine but vanished after reboot. Do you use some kind of immutable image, e.g. a pxe boot? Otherwise, maybe there was a kernel uname problem, please post the output of
ls -l /lib/modules

generix, thank you for the prompt response, I appreciate that.
image

image
Here is the screenshot with the output of
ls -l/lib/modules
When it comes to pxe boot - I guess it is not enabled, but, anyway, I will be able to check it a bit later.

The driver was only installed for the kernel
3.10.0-957.el7.x86_64
but you booted into
3.10.0-957.12.2.el7.x86_64
so it has no driver installed. Please use dkms to install the driver for different kernel versions.

1 Like

Thanks a lot, generix!
You was right. I’ve re-installed a driver using actual kernel libs. And after that a driver loads without any problems on start-up.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.