vCenter doesn't show ID for device backing from Datastores after decommissioning unrelated SAN LUN's.

Products

• VMware vCenter Server
• VMware vSphere ESXi 8.0
• VMware vSphere ESXi

Issue/Introduction

Symptoms: 

• When you try to view the device backing for a particular datastore, the list appears empty:

 

 

• You see that the hostd agent is unresponsive in the vobd.log on the ESXi host:
  
  
  2024-10-10T19:19:13.861Z In(14) vobd[2098339]:  An event (esx.problem.vmfs.ats.support.lost) could not be sent immediately to hostd; queueing for retry.
  
  
  

• From the vmkernel.log on the ESXi host, you may see SCSI H:0x1 errors ("no_connects") being reported:
  
  

2025-10-17T05:14:55.042Z In(182) vmkernel: cpu1:2097347)NMP: nmp_ThrottleLogForDevice:3893: Cmd 0x12 (0x45bc549ecc40, 0) to dev "naa.XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX" on path "vmhba4:C0:T9:L28" Failed:
2025-10-17T05:14:55.042Z In(182) vmkernel: cpu1:2097347)NMP: nmp_ThrottleLogForDevice:3898: H:0x1 D:0x0 P:0x0 . Act:NONE. cmdId.initiator=0x453c07f9bb18 CmdSN 0x0

 

• You were decommissioning other unrelated SAN LUN's

Environment

• VMware ESXi (All versions) 

Cause

• The LUN's may still have active I/O to them from things such as scratch partition, host log files, Cluster Services VM's, HA Datastore Heartbeating, etc., so when the LUN is unexpectedly removed, the hosts continue to try to send I/O, resulting the SCSI H:0x1 errors ("no_connect" errors) indicating the LUN can no longer be reached. 
• The ESXi host's hostd agent will subsequently become unresponsive due to the excessive outstanding I/O.

Resolution

• Please ensure that you follow the proper unmount/detach procedure for each of the SAN LUN's being decommissioned.  See KB for more information: Detach a LUN device from ESXi hosts.  
• A reboot of the host will resolve this, as it will no longer discover the LUN once it has been unmapped from the array side and thus the host will no longer try to use it for any system processes, logging, etc.