ESXi Host Unresponsiveness and vSAN Latency due to Intermittent Power Supply Unit (PSU) Failure.

Products

• VMware vSAN

Issue/Introduction

An ESXi host in a vSAN cluster becomes unresponsive, causing significant performance degradation across the cluster.  The following symptoms are observed:

• vCenter Connectivity: Host connection state changes to NO_RESPONSE in vpxd.log.

         2026-03-02T13:39:58.490Z warning vpxd[07002] [Originator@6876 sub=MoHost] host host-XX,XXXXX-XXXX-XXXX]connection state changed to NO_RESPONSE.

• Resource Starvation: vmkernel.log shows constant memory admission failures for various paths, including nsx-datapath-ctrs and backup.sh.

          Admission failure in path: host/vim/vmvisor/nsx-datapath-ctrs:python3
          Admission failure in path: host/vim/vmvisor/backup.sh

 

Environment

VMware vSAN 8.0U3

Cause

Hardware diagnostics identified that the Power Supply Unit (PSU)  intermittently generated errors. This hardware-level fault manifested at the hypervisor level as severe CPU and memory starvation, preventing the ESXi host from processing vSAN data and eventually leading to a complete hang of the management agents and data path.

Resolution

Replacing the failing hardware component resolves the resource starvation and host unresponsiveness.

1. Hardware Diagnostics: Run comprehensive hardware vendor diagnostics to confirm PSU health.
2. Hardware Replacement: Contact the hardware vendor to replace the faulty Power Supply Unit (PSU).