The prometheus-kube-prometheus-stack-prometheus-0 pod encountered a failure loop, halting the VCFA upgrade validation

Products

• VCF Automation

Issue/Introduction

- The VCFA upgrade precheck halted after detecting that mandatory monitoring services were unavailable

- Pod prometheus-kube-prometheus-stack-prometheus-0 in namespace vmsp-platform continuously crashes with CrashLoopBackOff

- Verification confirmed that storage allocation was sufficient and non-contributing

kubectl get pvc -n vmsp-platform | grep prometheus

Environment

VMware Cloud Foundation Automation 9.1

Cause

Prometheus pod exceeded its configured 1Gi memory ceiling during execution, triggering an operating system Out-Of-Memory (OOM) termination.

Resolution

1.Update the Prometheus resource definition in the vmsp-platform namespace to increase the memory limit from 1Gi to 8Gi:

kubectl patch prometheus kube-prometheus-stack-prometheus -n vmsp-platform --type merge -p '{"spec":{"resources":{"limits":{"memory":"8Gi"}}}}'

2.Confirm the pod successfully transitions to a Running state.

3.Rerun the VCFA upgrade precheck.