vIDB deployment fails using Fleet Management at Stage 4

Products

• VCF Operations

Issue/Introduction

The vIDB deployment fails at Stage 4 using Fleet Management with an error indicating a failure to create the services platform cluster. The specific error logs point to a failure to pull a Kubernetes image (cluster-api-installer) from the local registry due to a "context canceled" error, suggesting network connectivity or timeout issues during image retrieval.

 

/var/log/vrlcm/vmsp_bootstrap_###.log:

Failed to pull image "registry.vmsp-platform.svc.cluster.local:5000/images/cluster-api-installer:v1.0.7": rpc error: code = Canceled desc = failed to pull and unpack image "registry.vmsp-platform.svc.cluster.local:5000/images/cluster-api-installer:v1.0.7": failed to copy: httpReadSeeker: failed open: failed to do request: Get "http://127.0.0.1:30000/v2/images/cluster-api-installer/manifests/sha256:###?ns=registry.vmsp-platform.svc.cluster.local%!A(MISSING)5000": context canceled

Environment

VCF 9.0.x.0

Cause

A network policy that blocks unknown/new VM MAC addresses prevented the newly deployed vIDB VMs (which receive new MACs at deployment) from accessing the image registry.

Resolution

Allow newly assigned vIDB VM MAC addresses in the network policy so the VMs can reach the image registry during deployment.
NOTE: This will need to be done very quickly before the vIDB deployment process gets to the kube repo step

• Start the vIDB deployment and monitor for the VM power‑on event.
• As soon as the VMs are created, capture their assigned MAC addresses (for example, from your hypervisor/console).
• Register/allow‑list these MAC addresses in the control enforcing MAC‑based blocking (network policy/firewall/NAC).
• Continue the deployment and verify images pull successfully and the process completes without the timeout issue.