Restore VMware Guest files from Intellisnap backup with data on the secondary cluster (Snapvault)

@Gilles SCHMIDT , I was able to find something in our internal case history for this issue:

Please find the below Additional settings that you can update on the VSA proxy which will control the timeout for Vcenter API.

Right-click the VSA proxy>>Properties>>Advanced>>Additional settings.

nViSKTimeout

The default value as mentioned below is 10 minutes

You can change it to as below to 30 minutes. 

If that doesn’t resolve, we should look at the CVMA.log and snapmanager.log files for more info.

Hello Mike,

Issue is still present with additional settings but after parsing the log files, I have discovered it’s due the Ontap configuration. There was no export policy associated to SVM root volume to mount the volume (snapshot) on ESX. After modifcation, I was able to restore guest files. It’s performed in a lab with 2 NFS volumes as source and destination.

On customer site :

Primary storage Netapp : datastore (snapimirror source) is a FC LUN.

Destination storage Netapp : nfs volume (snapmirror destination).

My question is : with Commserve, can I restore data from a snapmirror destination as nfs volume (secondary storage for auxiliary copy) if my primary is an FC LUN

Many thx for your help

Best regards,

Gilles

Hi @Gilles SCHMIDT, I do not believe that will work - Commvault is going to try use the same protocol. There are some extra phases when protecting an NFS datastore during backup (like recording the NFS security configurations) that allows it to successfully mount the share on the remote side. 

File restores are a little different since by default we can now read the files directly from the snap mounted to the Media Agent rather than having to put the datastore back into vCenter. Obviously will full restores, you have to put the datastore back into VC/ESX to perform the restore from the snap copy.

Hi, I had a call with Commvault support. I had to check the communication between Netapp, MA, FREL and Esx is working well and transmit the FREL logs.

Communication on Ethernet is OK. After testing a browsing of guest files, I still received the error “failed to mount; Unrecognized error 30”. Nothing appears in logs the test period.

Is someone of you can explain where and how the volume(datastore) is mounting ?

Thx in advance,

Gilles

@Gilles SCHMIDT, can you share the case number so I can follow up and see their notes?

Hi Mike, here is the case number : 210715-453

Thanks for your help,

Gilles

Marking the above as the correct answer based on the case being archived after requesting a local proxy install.

If you are able to continue on this thread (if the solution presented didn’t work), update this thread and we will respond!

No worries, @Gilles SCHMIDT !  thanks for coming back and sharing with us.  I marked your reply as the Best Answer accordingly!

Thx Mike, have a nice day and Week-end

I would like to add to this thread in case someone else has similar problem.

We had the exact same problem. Error 30.

Looking in netapp audit log it states that snapclone cannot be created due to aggregate not found:

Error: Aggregate XXXXX not found. Reason: entry doesn't exist.

Solution:

Aggregate needs to be added to the SVM, and in our case it wasn´t. We have 3 SVM running in the cluster

vserver add-aggregates -vserver svm1 -aggregates aggr1,aggr2

After they were added I was able to browse and mount the snapclone to the MA via iSCSI