You can allow an OpenStack instance to use a NVIDIA GRID vGPU device on its ESXi host.

Note:
  • Only one vGPU is supported per OpenStack instance.
  • The same vGPU profile is used for all OpenStack instances.

Prerequisites

Verify that the driver for your vGPU device is installed on the ESXi host.

Procedure

  1. Log in to the OpenStack Management Server as viouser.
  2. If your deployment is not using a custom.yml file, copy the template custom.yml file to the /opt/vmware/vio/custom directory. 
    sudo mkdir -p /opt/vmware/vio/custom
sudo cp /var/lib/vio/ansible/custom/custom.yml.sample /opt/vmware/vio/custom/custom.yml
  3. Open the /opt/vmware/vio/custom/custom.yml file in a text editor.
  4. Uncomment the nova_gpu_profile parameter and set its value to the vGPU profile that you want to use.
    For example: 
    nova_gpu_profile: grid_p100-4a
  5. Uncomment the nova_profile_fb_size_kb parameter and set its value to the size of the vGPU frame buffer in kilobytes.
  6. Deploy the updated configuration. 
    sudo viocli deployment configure

    Deploying the configuration briefly interrupts OpenStack services.

  7. Log in to the VMware Integrated OpenStack dashboard as a cloud administrator.
  8. Select the admin project from the drop-down menu in the title bar.
  9. Select Admin > Compute > Flavors.
  10. Create a new flavor or choose an existing flavor to use for NVIDIA GRID vGPU passthrough.
  11. Select Update Metadata next to the flavor that you want to use.
  12. In the Available Metadata pane, expand VMware Driver Options for Flavors and click the Add (plus sign) icon next to VMware vGPU.
  13. Set the value of vmware:vgpu to 1 and click Save.

What to do next

You can now configure an instance to use an NVIDIA GRID vGPU by configuring the instance with the flavor that you modified in this procedure.