Is your feature request related to a problem? If so, Please describe.
When deploying dive/viame-web in NOAA GCP space, ie on a hardened VM, I've run into issues installing the GPU drivers via the nvidia.nvidia_driver role. Currently in the DIVE ansible playbook, there is no variable that can control whether the nvidia.nvidia_driver role is executed or not. My current solution is to comment out this line in the ansible yml files, before running the playbook.
Describe the solution you'd like
Updating the ansible playbook such that the user can provide a variable to not install the nvidia.nvidia_driver role. It feels reasonable to simply add when: run_worker_container or run_server or run_viame_cli | bool to the nvidia.nvidia_driver role?
Describe alternatives you've considered
The current workflow is functional, but indirect.
Additional context
N/A
Is your feature request related to a problem? If so, Please describe.
When deploying dive/viame-web in NOAA GCP space, ie on a hardened VM, I've run into issues installing the GPU drivers via the nvidia.nvidia_driver role. Currently in the DIVE ansible playbook, there is no variable that can control whether the nvidia.nvidia_driver role is executed or not. My current solution is to comment out this line in the ansible yml files, before running the playbook.
Describe the solution you'd like
Updating the ansible playbook such that the user can provide a variable to not install the nvidia.nvidia_driver role. It feels reasonable to simply add
when: run_worker_container or run_server or run_viame_cli | boolto the nvidia.nvidia_driver role?Describe alternatives you've considered
The current workflow is functional, but indirect.
Additional context
N/A