Error: "need 2 net devices, but found only 1 net devices" during Supervisor Control Plane VM deployment

Products

• VMware vSphere Kubernetes Service

Issue/Introduction

• During the deployment or lifecycle management of a Supervisor Cluster, a Supervisor Control Plane VM stalls in the "configuring workload network" phase.
• The affected Supervisor Control Plane shows the following error in the Supervisor management page of the vCenter UI:
  Need 2 net devices, but found only 1 net devices
• The /var/log/cloud-init.log file reports the following error in the affected Supervisor Control Plane:
  YYYY-MM-DD hh:mm:ss,### - performance.py[DEBUG]: cloud-init stage: 'init-network' took 0.## seconds
YYYY-MM-DD hh:mm:ss,### - handlers.py[DEBUG]: finish: init-network: FAIL: searching for network datasources
• The /var/log/vmware/wcp/wcpsvc.log file in the vCenter Server reports the following errors:
  YYYY-MM-DDThh:mm:ss.###Z WARNING wcp #### [vc@###] [kubelifecycle/controller.go:##] [opID=####-####-####-####] Unable to configure agent in supervisor ####-####-####-####. Err A general system error occurred. Error message: Need 2 net devices, but found only 1 net devices..
YYYY-MM-DDThh:mm:ss.###Z DEBUG wcp ####[vc@###] [kubelifecycle/upgrade_controller.go:##] [opID=upgrade-####-####-####-####] Supervisor ####-####-####-#### is not in READY phase, retry until it's ready
YYYY-MM-DDThh:mm:ss.###Z DEBUG wcp ####[vc@###] [kubelifecycle/upgrade_controller.go:##] [opID=upgrade-####-####-####-####] upgrade controller retrying
YYYY-MM-DDThh:mm:ss.###Z INFO wcp ####[vc@###] [kubelifecycle/vmc_alert_logging.go:##] [opID=upgrade-####-####-####-####] Upgrade for Supervisor ####-####-####-#### has taken ##### seconds; Upgrade has taken longer than 2 hours
• The second NIC fails to attach to the affected Supervisor Control Plane VM.

Environment

• VMware vSphere Kubernetes Service
• VMware Cloud Foundation (VCF) 9.x

Cause

The ESX Agent Manager (EAM) and VUM lifecycle hooks are out of sync with the current VM state. The deployment process expects a two-NIC configuration, but the lifecycle management process stalls, preventing the attachment of the second network interface.

Resolution

To resolve this inconsistency, manually re-trigger the VUM lifecycle task to synchronize the state and apply the required network configuration:

1. Navigate to the cluster hosting the Supervisor VMs in the vCenter UI to identify the Cluster ID (e.g., domain-c##).
2. Confirm the Solution ID by executing the following command:
   dcli com vmware esx settings clusters vms solutions list --cluster <cluster_id>
   
   
3. Re-trigger the VUM task to force the synchronization of the network configuration:
   dcli com vmware esx settings clusters vms solutions apply-task --cluster <cluster_id> --cluster-solutions-solutions <solution_id>
   
   
4. Verify the configuration in the vCenter Tasks view. Once the task completes, the second NIC attaches and the deployment proceeds.

Additional Information

For further assistance, contact Broadcom Support.