High Availability fails when same NFS share is mounted multiple times using different IP address

Products

• VMware vSphere ESXi

Issue/Introduction

This article provides information for customers that may accidentally mount the same NFS share multiple times using different IP addresses.

Symptoms:

• High Availability (HA) fails when same NFS share is mounted multiple times using different IP address.
• In the /var/log/vmkwarning.log file of the ESXi host, you see entries similar to:
  
  YYYY-MM-DDTHH:MM:SS.Z cpu24:3705703)ALERT: BC: 3118: File protectedlist closed with dirty buffers. Possible data loss.
  YYYY-MM-DDTHH:MM:SS.Z cpu24:3705703)WARNING: NFSLock: 4379: Unable to remove expired or lost primary lockfile .lck-04b04b0000000000
  YYYY-MM-DDTHH:MM:SS.Z cpu17:3705695)ALERT: BC: 3118: File protectedlist closed with dirty buffers. Possible data loss.
  YYYY-MM-DDTHH:MM:SS.Z cpu17:3705695)WARNING: NFSLock: 4379: Unable to remove expired or lost primary lockfile .lck-04b04b0000000000
  
  Note: The preceding log excerpts are only examples. Date, time, and environmental variables may vary depending on your environment.

Environment

VMware vSphere ESXi 7.0.0
VMware vSphere ESXi 6.7
VMware vSphere ESXi 6.5

Cause

This issue occurs when there are multiple datastores on an ESXi host, the High Availability (HA) application assumes there are separate filesystems.

It tries to create private files in each datastore for its own operation. If some of them are NFS datastores mounted from the same remote NFS share, the HA application runs into problems because its operating on the same files through multiple NFS datastores.

Resolution

This is by design.

VMware recommends to not mount the same remote NFS share multiple times using different IP addresses in a High Availability (HA) cluster.