After manual mounting from one site to another, the datastore is not visible.

Products

• VMware vSphere ESXi

Issue/Introduction

Symptoms:

This article helps you identify problems related to mounting datastore manually from one vCenter to another.

• ESXi host cannot see datastore mounted manually.

• The datastores at the recovery(DR) site are unmounted and then mounted on the production(Prod) site hosts, after which they become invisible.

• From logs the LUN appears as snapshot.

  /var/run/log/VMkernel.log

  ####-##-##T##:##:##.893Z cpu41:2105470)BC: 414: read from naa.################################ (def5 24 c1 3 d3c119c0 4307 3 0 0 0 0 0 0 0) 69632 bytes failed: SCSI LUN is in snapshot state

• The adapter/storage rescan has completed successfully, but the datastore is still missing.

• Running below command shows no results.

  esxcli storage vmfs snapshot list

Environment

VMware vSphere ESXi 8.x

Cause

• This issue occurs if an Storage Recovery activity is carried out without performing actual Recovery steps.

• The storage from Recovery site vCenter is manually disconnected and reconnected on the Production site vcenter resulted in the Storage to remain in write mode instead of Read/Write mode. Consequently, the datastore was not accessible even after it was mounted.

From the logs /var/run/log/vmkernel.log,

####-##-##T##:##:##.892Z cpu41:2105470)Partition: 437: Failed read for "naa.################################": I/O error
####-##-##T##:##:##.892Z cpu41:2105470)Partition: 1123: Failed to read protective mbr on "naa.################################" : I/O error
####-##-##T##:##:##.892Z cpu41:2105470)WARNING: Partition: 1289: Partition table read from device naa.################################ failed: I/O error
####-##-##T##:##:##.893Z cpu30:2097971)ScsiDeviceIO: 4167: Cmd(0x45d916888948) 0x28, CmdSN 0x1 from world 2105470 to dev "naa.################################" failed H:0x0 D:0x2 P:0x2
####-##-##T##:##:##.893Z cpu41:2105470)BC: 414: read from naa.################################ (def5 24 c1 3 d3c119c0 4307 3 0 0 0 0 0 0 0) 69632 bytes failed: SCSI LUN is in snapshot state

####-##-##T##:##:##.930Z cpu29:2097971)NMP: nmp_ThrottleLogForDevice:3798: last error status from device naa.################################ repeated 40 times
####-##-##T##:##:##.930Z cpu29:2097971)ScsiDeviceIO: 4167: Cmd(0x45b91a018648) 0x28, CmdSN 0x11 from world 2102786 to dev "naa.################################" failed H:0x0 D:0x2 P:0x2

Resolution

• As the proper DR procedure was not adhered to and the Recovery Site's Storage was mounted manually with its mode remaining in write rather than RW, the Storage Vendor is required to be engaged.

• Datastore to be mounted as a snapshot in order to change the state to RW, if mounted Manually the state will remain in Write status. 

• Below command to run once the datastores are mounted as snapshot,

  To list the volume detected as snapshot,

  esxcli storage vmfs snapshot list

  
  Volume Name: VMFS_1
VMFS UUID:########-########-####-############
Can mount: true
Reason for un-mountability:
Can resignature: true
Reason for non-resignaturability:
Unresolved Extent Count: 1

• Resignature of datastore snapshot needs to be carried out using below command to mount the datastore,

  esxcli storage vmfs snapshot resignature -l label

  For complete steps refer - Troubleshooting LUNs detected as snapshot LUNs in vSphere