Long snapshot stun time observed for VMs running datastore backed by NVMe-TCP (Pure Storage)

Products

• VMware vSphere ESXi

Issue/Introduction

Symptoms :

• Virtual machines (VMs) running on datastores backed by NVMe-TCP (Pure Storage) may experience high stun times during snapshot operations. This behavior is typically observed when backup software initiates a snapshot or when a manual snapshot operation is performed.
  
  Example : 
  
  Stunned Time :
  2025-10-28T07:42:57.954Z In(05) vcpu-0 ########-auto-#####-##:########-##-##-#### CPT: vm was stunned for 2584922 us
  2025-10-28T07:43:07.614Z In(05) vcpu-0 ########-auto-#####-##:########-##-##-#### CPT: vm was stunned for 266726 us
  
  
  
  

Environment

• Storage is connected via NVMe-TCP protocol (Pure Storage)
  
  
• VMware vSphere ESXi 8.x

 

Cause

This issue occurs when VMs are hosted on datastores connected through NVMe-TCP using Pure Storage Purity//FA version 6.5.8.

A condition in this version results in prolonged VM stun times during snapshot creation in VMware environment. 

Resolution

The issue has been resolved in Pure Storage Purity//FA version 6.9.1. 

With this release, VMs running on datastores connected via the NVMe-TCP protocol should no longer experience extended stun times during snapshot operations. 

Please contact Pure Storage for further assistance on Pure Storage upgrade. 

Additional Information

Purity Simplified Release Lines