Open paveldournov opened 1 year ago
This error occurs when we create separate VPC network. We have moved to using default network instead of creating a VPC network every time. with that change, this error is not happening any more.
Please use the latest or lkg image of us-central1-docker.pkg.dev/supercomputer-testing/cluster-provision-repo/cluster-provision-image.
Still am getting the error on a newly built image. The tool leaves the VMs running after failing the Destroy.
module.compute-vm-1.google_compute_instance_group_manager.mig: Refreshing state... [id=projects/test-vms/zones/us-central1-f/instanceGroupManagers/aicluster-mig] ╷ │ Error: Unsupported attribute │ │ on main.tf line 42, in module "compute-vm-1": │ 42: subnetwork_self_link = module.network1.subnetwork_self_link │ ├──────────────── │ │ module.network1 is object with 4 attributes │ │ This object does not have an attribute named "subnetwork_self_link". ╵
Added more debugging information with PR https://github.com/GoogleCloudPlatform/ai-infra-cluster-provisioning/pull/39
╷ │ Error: Unsupported attribute │ │ on main.tf line 42, in module "compute-vm-1": │ 42: subnetwork_self_link = module.network1.subnetwork_self_link │ ├──────────────── │ │ module.network1 is object with 4 attributes │ │ This object does not have an attribute named "subnetwork_self_link". ╵