CODEWITHSUNDEEP
google-cloud-platformnvidia
Eduardo Maciel
Eduardo Maciel

Reputation: 31

GCloud VM instance unloaded Nvidia drivers after reboot

I've been working with a VM instance on GCP for a while now and I had to reboot it because I had issues logging in through SSH. I used the "Reset" option in the VM Instances manager on the GCP website and the Nvidia drivers seem to be down after that.

I tried running nvidia-smi and got the following:

NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.

I don't know how to set it back up or why did this happen in the first place. Because of this I haven't been able to get my Docker containers back up.

Upvotes: 1

Views: 922

Answers (1)

Mahboob
Mahboob

Reputation: 1975

There is a PIT-1, PIT-2 and noticed that there are some WA for Ubuntu, you can try those; otherwise its seems to me that the issue you are having at the moment is intended behavior.

These are the steps to successfully install the drivers on CentOS 7:

  1. Run sudo yum -y update kernel
  2. Run sudo yum -y update
  3. Stop and Start the instance
  4. Run sudo yum install kernel-devel kernel-headers -y
  5. Follow the documentation starting at step 3 1.

If you run "nvidia-smi" after following the instructions and stopping and starting the instance, you will be able to see the proper output as mentioned in the doc.

Upvotes: 1

Related Questions