After Shrinking LUN — Datastore Cannot Be Deleted : Error An error occurred during host configurations Operation failed Unable to find a partition with the name naa.00

Products

• VMware vSphere ESXi
• VMware vCenter Server

Issue/Introduction

 ESXi hosts connected over iSCSI experience an issue after reducing the size of a VMFS datastore from 35 TB to 20 TB directly on the storage array. Post-resize, the datastore displayed 0 TB capacity in vCenter, and deletion attempts failed.

An error occurred during host configuration: Operation failed, diagnostics report: Unable to find a partition with the name found: VSI node _storage_scsifw_devices_partitions_info)

Environment

 

• vCenter Server 

• ESXi hosts 

 

Cause

The issue was triggered by a direct LUN resize operation (shrink) performed on the Datastore without properly detaching or unmounting the VMFS datastore from the ESXi hosts.

This resulted in an inconsistent state:

• Two ESXi hosts showed the LUN as detached and unconsumed

• One ESXi host still listed the datastore as active, holding stale metadata

VMware does not support shrinking VMFS datastores from the backend without coordination from the vSphere layer. Please see Correct procedure to reduce the size of a VMFS datastore

Resolution

 

• Identify the ESXi host that still reports the datastore as active (serv-esxi03 in this case).

• Migrate all VMs from the affected host to another host in the cluster.

• Place the host into Maintenance Mode via the vSphere Client. (Optional if decided to reboot the host instead)

• Restart all services on the host or reboot the host to clear stale datastore references:

  • To restart services:

    services.sh restart

  • Or perform a full reboot of the host.

• After the host is back online, exit Maintenance Mode.

• Confirm that the stale datastore reference is cleared on all hosts and vCenter.