vSphere - NetWorker SQL backups fail with error: "Unable to find VSS metadata files in directory"

-
In this post I want analyze a curious behavior of NetWorker SQL backup in vSphere enviroment:

Symptoms

An examination of the MSVMAppAgent logs on the SQL server show the following errors:

C:\Program Files\DPSAPPS\MSVMAPPAGENT\logs\msvmcatsnap.log


171634:msvmagentcatsnap:

===Command Line Arguments===: 

C:\Program Files\DPSAPPS\MSVMAPPAGENT\bin\msvmagentcatsnap.exe -i -Extracted VSS metadata file 'C:\Program Files (x86)\EMC\vProxy FLR Agent\tmp\695ccb0f-b0f3-43a5-942f-e0c83fcafe26\vm-180966-vssManifest.zip' to directory 'C:\Program Files\DPSAPPS\MSVMAPPAGENT\tmp\VSSMetadata.2021-07-30_09-08-58'168886:

msvmagentcatsnap: Unable to find VSS metadata files in directory C:\Program Files\DPSAPPS\MSVMAPPAGENT\tmp\VSSMetadata.2021-07-30_09-08-58.102333:msvmagentcatsnap:

 Exiting with failure.

Cause

When VMware takes a snapshot of a Windows VM with the Quiesce option enabled it uses the installed VMware tools to query theMicrosoft Volume Shadow Copy service.

Then creates a file called VM_name-vss_manifest#.zip in the VM's datastore foldercontaining the VSS configuration metadata.

That file is copied into the VM during the vProxy based SQL backup and parsed for the specific configuration details. If that file is notpresent or not complete the backup will fail with the error noted above.

You can manually verify the contents of this file using the procedure below:


1. Take a Snapshot of the SQL server in VMware and specify the Quiesce option:


2. From the Data Store Browser download the "VM-Name-vss_manifest##.zip file and examine the contents.
An intact example of the file contents similar to those shown below:

An example of an incomplete version of the file is shown below:


Resolution

After a deep analysis the error was due to the missing VM parameter disk.EnableUUID = “TRUE” inside the .VMX.
The curious fact is that happen only with Windows Server 2016 VMs and is described on this KB https://kb.vmware.com/s/article/52815.

If you add manually disk.EnableUUID = “TRUE” parameter in the VM configuration the error will be resolved and the backup job compete successfully.

Thanks for reading!

Commenti

Posta un commento

Post popolari in questo blog

Aria Operations unable to edit credentials

-
Immagine
  After upgrade Aria Operation from 8.14 to 8.16 one of my customers reporting me that can't manage anymore all credentials. The two options under 3-dots "Edit" and "Delete" was grayed out. I found the solution under the release notes; credentials now require an owner and after assign ownership, are editable once again. Below useful links and detailed snapshot: https://docs.vmware.com/en/VMware-Aria-Operations/8.14/rn/vmware-aria-operations-814-release-notes/index.html https://docs.vmware.com/en/VMware-Aria-Operations/8.14/Configuring-Operations/GUID-0D87C202-6F7D-4A7D-A69D-953AC382E37B.html Thanks for reading!
Continua a leggere

Aria Operations USB controller report for Critical Severity VMSA-2024-0006

-
Immagine
Yesterday VMware released VMSA 2024 0006 that impact all VMs that have configured an USB controller. Matrix response give us two choice: install ESXi patch or apply a workarounds that consist to remove the USB controller from a VMware ESXi virtual machine. If you decide to use the second option I created a report on Aria Operation to determinate which VMs are impacted.  Below few simple steps to setup the report: 1 - Firs of all you must enable "Disconnected USB controller" properties, in Configure -> Policies -> Policy definitions: 2 - Choose your "Active" policy and click on "edit policy" 3 - Click on "Metrics and Properties" 4 - Search "USB" in virtual machine object type and activate the properties 5 - Create or edit a view adding the "Disconnected USB controller" 6 - Create or edit a report based on the previous view 7 - Filter the column by "false" Now you have the VMs list affected where you must to remo...
Continua a leggere

Aria Operations for Logs - Increase datastore alert

-
Immagine
  Here a quick and easy alert configuration in Aria Operations for Logs helpful to know when, how much and who increase a datastore size. In query section insert the two text conditions that contain the key words "user:" and "increased in capacity". Don't forget to put "any" to ensure to match both conditions. Lastly put your email address in "Trigger condition" to notify yourself. Thanks for reading!
Continua a leggere