Datastore unmount fails with "Select at least one object to continue" error in vSphere Client

Products

• VMware vCenter Server

Issue/Introduction

• • When attempting to unmount a datastore from the vCenter vSphere Client, the operation cannot be completed.

  • The "Unmount Datastore" wizard displays the error message: Select at least one object to continue.

  • The wizard shows COMPATIBLE (0 HOSTS) and explicitly reports No compatible hosts found, preventing the user from proceeding with the unmount task.

  • The datastore appears to be "empty" or "disconnected" from user virtual machines.

  • This frequently occurs after a storage array power failure or EOL hardware decommissioning where the underlying LUN was removed from the storage array before being properly unmounted in vCenter.

Environment

VMware vCenter Server 8.x
VMware ESXI 8.x
Storage Type: VMFS datastore

Cause

vCenter prohibits the unmount or deletion of datastores that are actively referenced by cluster services or hosts. Even if all user VMs are migrated off the datastore, this issue occurs when vCenter's inventory is out of sync with the actual ESXi hosts, or when hidden system-level dependencies hold an active lock on the volume.

Specific causes include:

• vSphere Cluster Services (vCLS) VM placement locks: Hidden system VMs are residing on the datastore.

• Stale vCenter database entries due to improper LUN detachment (APD/PDL): The underlying LUN was unmapped or dropped from the storage array before a clean unmount was performed in vCenter.

• Registered VM Templates or stale ISO/CD-ROM mount points: Templates do not always appear in the standard VM view, and ISOs/Floppies originating from the datastore may still be mounted to other VMs.

• System Configurations: The datastore is being used for vSphere HA heartbeating, Storage I/O Control (SIOC), or persistent scratch logging.

• Host Connectivity: One or more ESXi hosts in the cluster are in a "Disconnected" or "Not Responding" state.

Resolution

Please follow the steps below to identify and release the locks on the datastore.

Phase 1: Clear Stale Inventory Data

Often, a simple rescan will force vCenter to update its inventory, clearing phantom locks.

1. Right-click your Cluster or Datacenter and select Storage > Rescan Storage.

2. Check the boxes for Scan for new Storage Devices and Scan for new VMFS Volumes.

3. Once the task completes, refresh the vSphere Client browser and attempt the unmount again.

If the rescan does not resolve the issue, proceed to Phase 2.

Phase 2: Remove vCLS VMs (Retreat Mode)

If the datastore is still locked, temporarily remove the hidden vCLS system VMs from the cluster.

1. Navigate to Cluster > Configure > vSphere Cluster Services > General.

2. Click Edit vCLS Mode and select Retreat Mode.

3. Wait for the vCLS VMs to be powered off and deleted from the cluster.

4. Attempt to unmount the datastore again.

   • Note: Once the datastore is successfully unmounted, remember to return the vCLS mode to System Managed.

Phase 3: Verify System Configurations, VM Templates, and Active Mounts

Check for the following configurations that may be locking the datastore:

• VM Templates: Check for VM Templates registered to the datastore. These do not always show in the standard "VMs" view and must be unregistered or moved.

• ISO/Floppy Mounts: Ensure no powered-on or powered-off VMs have ISOs or Floppy drives connected to files residing on this datastore.

• HA Heartbeating: Go to Cluster > Configure > vSphere Availability > Edit > Heartbeat Datastores. Ensure the target datastore is not selected for datastore heartbeating.

• Host Connectivity: Verify all hosts in the cluster show as "Connected." If a host is disconnected, vCenter cannot route the unmount command properly.

Phase 4: Force Unmount via ESXi CLI (Advanced)

If the UI remains unresponsive or displays 0 compatible hosts despite the above checks, unmount the datastore directly via the ESXi command line.

1. Enable SSH and log into the ESXi host that previously had access to the datastore.

2. List the mounted volumes to verify the datastore's presence:

   
   

    

   esxcli storage filesystem list
   

3. Unmount the datastore using its label (replace Datastore_Name with your actual datastore name):

   
   

    

   esxcli storage vmfs unmount -l Datastore_Name
   

4. Return to vCenter and perform a Storage Rescan.

Phase 5: Restart vCenter Services (Last Resort)

If all previous phases fail and the stale inventory persists (typically exhibiting as an unremovable ghost datastore), you may need to restart the vCenter Server Appliance (VCSA) services.

1. Connect to the VCSA VAMI or SSH into the appliance.

2. Reboot the VCSA or restart the vpxd services to force a fresh database sync with the ESXi hosts.

Additional Information

For more detailed instructions and specific error resolutions regarding datastore removal, refer to the following Broadcom Knowledge Base articles:

• • [321565] How to unmount a LUN or detach a datastore device from ESXi hosts

  • [311779] Error: "Device or Resource Busy" or "Device in use" when attempting to delete/unmount/disconnect a Datastore

  • [386291] Unmounting the datastore failed to remove from vCenter 'Not connected and accessible host is attached to this'

  • [392105] Unable to delete empty datastore