PPDM: vSphere reporting "Virtual machine Consolidation Needed" for VMs where consolidation task completes successfully

• PowerProtect Data Manager VADP VM backups in use
• "Exclude Swap files" enabled on VM Protection Policy
• vSphere consolidation task completes successfully yet vSphere still reporting "Virtual machine Consolidation Needed" for VMs
• No snapshots present in vSphere UI
• vm-00001.vmdk snapshot files present on VM

PPDM-274062

Details: Ensures VM Direct Engine (vProxy)'s VMDK Fuse driver closes any connection to .vmdk disks following VADP backup completion.
Affected versions: Potentially could impact all versions of PowerProtect Data Manager.
Fixed Version: 19.16 and 19.15.0-25 CHF

PPDM-252524

Details: VM Direct Engine (vProxy) vmdkfs process becomes defunct and holding a lock on the .vmdk fixed.
Affected Version: Potentially could impact all versions of PowerProtect Data Manager.
Fixed Version: 19.14.0-27, 19.15.0-25 CHF and above

Resolution:
Upgrade to PowerProtect Data Manager 19.15.0-25 or higher.

Workaround for both POWERPROTECT DATA MANAGER-274062 and POWERPROTECT DATA MANAGER-252524

1. Power off and on the VM Protection Engine (vProxy) using vSphere and wait 5-10 minutes before proceeding to step 2.

   Note: Restarting the Guest operating system does not always clear this issue

    

2. vSphere->Right-click the VM->Snapshots->Consolidate. Once complete, the vSphere alert "Virtual machine Consolidation Needed" should no longer remain.

To verify if POWERPROTECT DATA MANAGER-252524 is the issue affecting your environment.

1. Connect to VM Direct Engine (vProxy) over ssh

2. Run on vProxy to check for defunct process.

   # ps -ef | grep vmdkfs

   

   Note: To view the admin/root password for the VM Direct Engines, use the commands below run as root on the PowerProtect Data Manager. This lists all passwords for all VM Direct Engines, both VM, and NAS.
   source /opt/emc/vmdirect/unit/vmdirect.env  &&  /opt/emc/vmdirect/bin/vproxymgmt get -secret

    

3. Tcp connections can also be observed left in a "CLOSE_WAIT" status between the VM Direct Engine and the ESXi host. Run below on the VM Direct Engine over ssh

   # netstat -anp | grep -E ":902"

   

   Or
   # ss -at <-- for new versions of Suse Linux where netstat is no longer in use.