NSX Distributed Load Balancer reported as Degraded due to Supervisor Service Image Pull Failures
search cancel

NSX Distributed Load Balancer reported as Degraded due to Supervisor Service Image Pull Failures

book

Article ID: 437787

calendar_today

Updated On:

Products

VMware Cloud Foundation VMware NSX

Issue/Introduction

  • NSX Distributed Load Balancer (DLB) status shows as Degraded or "partially ready"  in NSX Manager UI - Networking - Load Balancer.
  • The associated Logical Switch Ports (LSP) are in a "Not Ready" state.

get load-balancer <LB_UUID> status

Status: partially ready
Ready LSP Count: 2
Not Ready LSP Count: 1
Not Ready LSP: <LSP_UUID>

  • Supervisor service pods (e.g., cci-ns-controller-manager) in Pending state from vSphere client - Inventory.
  • kubectl describe pod <Supervisor service pods> found failed reason: ErrImagePull.

Environment

VCF 9.0.X

Cause

Supervisor VM eth1 workload interface is blocked from accessing the Broadcom image registry (projects.packages.broadcom.com), preventing required container images from pulling.

Because the pod never starts, the VIF never connects to the switch, leaving the LSP in a Not Ready state.

Resolution

  • Identify the egress IP range for the Supervisor Workload Network.
  • Ensure the physical firewall allows traffic on Port 443 (HTTPS) from the Supervisor eth1 subnet to projects.packages.broadcom.com.
    Validate connectivity from a Supervisor control VM:
    curl -v --interface eth1 https://projects.packages.broadcom.com
  • Delete Pending pods (e.g., cci-ns-controller-manager) to force pod restart:
    kubectl delete pod <pod-name> -n <namespace>