Warning: "Virtual Machine Disk Consolidation Required" after recovery of a VM using vSphere Replication

Products

• VMware vSphere ESXi
• VMware Live Recovery
• VMware vSAN 8.x

Issue/Introduction

• After a successful recover using vSphere Replication, a VM throws the below warning on the vCenter server UI:

Virtual machine VCF-TEST-Replication1 disks consolidation is needed on XXXXXX.com in cluster YYYY.com in ha-datacenter.

• The issue only occurs when the VM is recovered with the target site running on vSAN ESA or vVOL
• It works as expected without any warning when the recovery is run with target site running on vSAN OSA, VMFS or NFS
• The below event is observed on the vCenter server:

Virtual machine disks consolidation is needed.    Warning    01/31/2025, 1:33:29 AM        VCF-TEST-Replication1      com.vmware.vc.VmDiskConsolidationNeeded

• The VM appears to be running on a snapshot disk on the vCenter server
• This snapshot is the image disk created during the recovery  operation:

Environment

• VMware vSphere Replication
• Target Site - VMware vSAN ESA 
• VMware vSphere ESXi

Cause

• This is a known issue effecting the recovery workflow using vSphere Replication to vSAN ESA datastores
• During recovery, multiple changes to disks are made to the "VMX" file 
• While recovering to a vSAN ESA environment, a missing reload of "VMX" causes a difference in information regarding disks between "hostd" and the "VMX":

• "hostd" has the below information for the VM configuration You will notice the HBR image disk as the file for the disk:

(vim.vm.device.VirtualDisk) {
dynamicType = <unset>,
dynamicProperty = (vmodl.DynamicProperty) [],
key = 2000,
deviceInfo = (vim.Description) {
   dynamicType = <unset>,
   dynamicProperty = (vmodl.DynamicProperty) [],
   label = 'Hard disk 1',
   summary = '94,371,840 KB'
},
backing = (vim.vm.device.VirtualDisk.FlatVer2BackingInfo) {
dynamicType = <unset>,
dynamicProperty = (vmodl.DynamicProperty) [],
fileName = '[XXXXX-vsan01] d6556267-cc4b-fd76-5407-XXXXXXXX/hbrimagedisk.RDID-3657f2b3-2a29-4adf-a77b-451e2ac85239.vmdk',

• While the VMX has the below configuration:

scsi0.virtualDev = "lsisas1068"
scsi0.present = "TRUE"
scsi0.pciSlotNumber = "160"
scsi0.sasWWID = "50 05 05 6b 21 37 fb a0"
sched.scsi0:0.shares = "normal"
sched.scsi0:0.throughputCap = "off"
scsi0:0.deviceType = "scsi-hardDisk"
scsi0:0.present = "TRUE"
scsi0:0.redo = ""
scsi0:0.fileName = "/vmfs/volumes/vsan:523182a14c114ef5-XXXXXXXXX/d6556267-cc4b-fd76-YYYYYYYY/VM-TEST-Replication1.vmdk"

 

• This causes "hostd" to post a "Consolidation Needed" event which is also passed on to the vCenter with the configuration that the "hostd" has
• The VM meanwhile is powered on using the base disk that the VMX configuration has

 

Resolution

• This will be fixed in the upcoming 9.0.3 release of vSphere Replication which will include another reload of the VMX at the end of he recovery workflow
• Once the fix is in place, "hostd" will update the correct values from the VMX during the reload operation on the VMX
• This will ensure that the correct information is passed on to the vCenter server and the information is refreshed, ensuring the warning to disappear

Workaround:

Option 1:

• Migrate or vMotion the VM to another ESXi host post recovery to reload the VMX

Option 2:

• Use the snapshot manager to consolidate the disks which will reload the VMX

Option 3:

• Reload the VMX from the command line
• Follow the below KB to reload the VM from the command line of the ESXi
  Reloading a VMX file