VCF Automation and Migration Service Engine upgrades failed during the precheck and staging phases.

Products

• VCF Automation

Issue/Introduction

- Prechecks failed for VCF automation and VCD migration on stage phase

- VCFA node status reported as SchedulingDisabled

- Numerous support-bundle-collect-metrics cronjobs remained in a Pending state

Environment

VCF Automation 9.1

Cause

Multiple support-bundle-collect-metrics cronjobs were stuck in pending, consuming system resources, then VCFA node entered SchedulingDisabled status due to VM restart and preventing new pods from starting, both the underlying resource shortage and node status caused the vmsp_stage_plugin and stage_vsp_cluster_plugin steps to time out or crash

Resolution

1.Delete all pending support-bundle-collect-metrics jobs from the VCFA node.

kubectl get jobs -n vmsp-platform | grep support-bundle-collect-metrics | awk '{print $1}' | xargs -I {} kubectl delete job {} -n vmsp-platform

2.Recovery the cluster to clear the SchedulingDisabled status refer to KB440862

3.Verify all VCFA pods return to a Running/Normal state.

4.Re-run the upgrade precheck and proceed with the VCF upgrade.