Unable to see datastore from the ESXi hosts.

Products

• VMware vSphere ESXi

Issue/Introduction

• You can see the paths are active for the datastores from the ESXi host CLI.
• From the vmkernel log you can see the below scsi sense code for the LUN or datastore.

yyyy-mm-dd:hh:mm:ss cpu##:2098117)WARNING: NMP: nmpCompleteRetryForPath:364: Retry cmd 0x28 (0x45d95528d5c8) to dev "naa.xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" failed on path "vmhba#:C#:T#:L##" H:0x0 D:0x2 P:0x0 Valid sense data: 0x2 0x4 0x3.

• The scsi sense code indicates the LUN is not ready.

Sense Key    [0x2]    NOT READY

Additional Sense Data    04/03    LOGICAL UNIT NOT READY, MANUAL INTERVENTION REQUIRED

OP Code    0x28    READ(10)

• From the Storage array GUI or CLI, the status of the LUN is showing as NVFail. In the below image the status of the LUN will show as NVFail instead of online.
  
  

Environment

• VMware vSphere ESXi 7.x
• VMware vSphere ESXi 8.x

Cause

• The SCSI sense code indicates that the LUN is either offline or unable to prepare for any operations. Therefore, this issue needs to be investigated manually on the storage array side.
• The READ(10) command with Operation Code (OP Code) 0x28 is part of the SCSI command set, which is commonly used for communicating with storage devices or LUN. In this case, the LUN is failing to respond to the READ request due to its current status.
• The LUN is showing an NVFail status, which indicates that NVRAM errors have been detected. NVFail can block database instances from accessing data until the system administrator manually verifies the data integrity or addresses the issue

 

Resolution

Please engage your storage vendor to investigate.

 

 

 

Additional Information

The NVFail issue on a LUN can occur across multiple types of storage arrays.

 

But find the ONTAP document for more details as example.