Epic Goal

• Make hibernation feature able to stop GCP VirtualMachines which have Local SSD(s) attached (e.g., a2-ultragpu-4g- NVIDIA A100 GPUs). 

Why is this important?

• clusters with GPU nodes becomes partially hibernated with all the nodes except GPU ones stopped - GPUs keeps being billed
• user don't have idea of it happening until they check on Google Cloud console or read the Hive ClusterDeployment status
• it affects both Managed (OSD) and self-managed clusters

Scenarios

1. Create a cluster with Hive on GCP
2. add a GPU worker node to the cluster - use flavor a2-ultragpu-4 for example
3. trigger cluster hibernation via Hive
4. check the VM status in GCP console
5. check the Hive Cluster Deployment conditions

Acceptance Criteria

• GPU nodes get hibernated as the other worker/master nodes
• Hive exposes a corresponding option to --discard-local-ssd (
  https://cloud.google.com/compute/docs/disks/local-ssd#stop_instance)
• 
  others TBD 

Dependencies (internal and external)

1. OCM has a dependency on Hive ?

Previous Work (Optional):

1. …

Open questions::

Done Checklist

• CI - CI is running, tests are automated and merged.
• Release Enablement <link to Feature Enablement Presentation>
• DEV - Upstream code and tests merged: <link to meaningful PR or GitHub Issue>
• DEV - Upstream documentation merged: <link to meaningful PR or GitHub Issue>
• DEV - Downstream build attached to advisory: <link to errata>
• QE - Test plans in Polarion: <link or reference to Polarion>
• QE - Automated tests merged: <link or reference to automated tests>
• DOC - Downstream documentation merged: <link to meaningful PR>