When I try to deploy this image to GCP, this is the error that pops up:
nvidia-gpu-cloud-pytorch-image-1-vm: {"ResourceType":"compute.v1.instance","ResourceErrorCode":"400","ResourceErrorMessage":{"code":400,"errors":[{"domain":"global","message":"e2 instances do not support onHostMaintenance=TERMINATE unless they are preemptible.","reason":"badRequest"}],"message":"e2 instances do not support onHostMaintenance=TERMINATE unless they are preemptible.","statusMessage":"Bad Request","requestPath":"https://compute.googleapis.com/compute/v1/projects/xxx/zones/xxx/instances","httpMethod":"POST"}}
It seems to be impossible to manulally change “On host maintenance” setting in this type of deployment. Can it be fixed?