Skip to main content

Hi,

I am in the process of configuring NetApp NAS backups using CIFS with SnapDiff, and I cannot find a clear documentation on how to configure that. What is unclear for me, is:

SnapDiff is just used for the Scan phase, right? So once we are in the Backup phase, we are just doing a CIFS--based backup?

  1. In which steps should we configure this? Create the NAS client first? Or configure the Array first? And at which point (and where in Command Center) should we detect/configure the SVM's?
  2. What permissions within NetApp does the user account need which is used for the SnapDiff feature?
  3. After Configuring the Array and the CIFS subclient, will Commvault automatically use SnapDiff? Or should we enable that somewhere?
  4. How does Commvault map the share to the NetApp Array and volume that needs to be used for SnapDiff? For instance:
    - The NetApp Array is configurated in Commvault as: netapp01-mgmt.company.lan
    - Clients connect to \\netapp01-svm1-cifs.company.lan\share01
    - We run the backup as: \\netapp01-svm1-backup.company.lan\share01
  5. How can we see if SnapDiff or traditional recursive Scan was used for a CIFS backup?

Sorry if these are stupid questions, but documentation seems to be scattered all over the place. No coherent explanation/overview. Or if there is, please point me towards it. :-)

Hello,

please find the answers below:

1) We should add the NAS client first and then add the array management entry for it. Once both have been added you can follow the doc below to detect/add svms: https://documentation.commvault.com/11.24/expert/35131_detecting_and_adding_storage_virtual_machines_on_netapp_arrays.html
2) There are no additional permissions that need to be added to the service account used as the flag to leverage snapdiff is sent using the credentials configured in the array management.
3) If intellisnap is enabled for the subclient in question all backup will use snapdiff by default as explained in the doc below. I cant say I have seen a non intellisnap backup that uses snapdiff but you can configure it with the basic config and see if it still leverages snapdiff.
https://documentation.commvault.com/11.24/essential/131717_faster_incremental_backups_for_nas_file_servers_using_netapp_snapdiff.html
4) Commvault uses the array access nodes defined in the subclient properties to access the snapshot taken by the snap backup over UNC \\Files\Volume\.snpshot\Snapshot_Taken_by_snap backup.
5) You can monitor the file scan logs on the access nodes to confirm snapdiff is in use. 

Again this is all assuming intellisnap is enabled. If it is not enabled and we want to leverage snapdiff, I would recommend opening a ticket with support so they can get confirmations on the steps.

Cheers
 

Hi @Harsh Desai , thanks for you reply.

We don’t want to use IntelliSnap, we just want to use the CIFS backup method and leverage SnapDiff just for speeding up the Scan pase. So maybe I will have to log a support ticket for that then. Just a few follow-up questions.

 

Regarding question 2. You say “There are no additional permissions that need to be added”. No additional as opposed to what? Is is compared to a regular IntelliSnap user? The customer has created a custom account and wants to know which permissions he needs to assign for SnapDiff to work. Can you point me to the Documentation page where the permissions are listed? Because I am having trouble finding them.

 

Regarding question 4. Maybe my question was not clear. I will try to rephrase. Let’s say she share is configured on NetApp as \\netapp01-svm1-cifs.company.lan\share01. But I want to connect to a dedicated backup LIF on the NetAppp, so I create a DNS alias and connect to \\netapp01-svm1-backup.company.lan\share01 in Commvault. How will Commvault know which Filer and which Volume to take a snapshot on? 

 

Regarding question 5. So it is not shown in the Command Center or Java GUI? Normally we can see what Scan method is used (Recursive/Change Journal/Optimized Scan). So it will not show “SnapDiff” or something there?

 

Thanks!

Hi @Patrick Dijkgraaf !  I spoke to a few different colleagues here who suggested opening a case as we’d likely need to engage development to assist.

In the meantime, I’ll contact our docs team about this to get the stuff you’re looking for documented fully!

Can you share the case number once created?

Thanks!

I spoke to our docs team who advised we have instructions from here:

https://documentation.commvault.com/11.24/essential/131717_faster_incremental_backups_for_nas_file_servers_using_netapp_snapdiff.html

They also asked me to get a list of items you are not seeing that you need to find and I can create a documentation MR in your name.

Thanks!

Hi @Mike Struening,

Thanks so far! I’ll gather the information and open a support ticket and report back here afterwards. Might take 1 or 2 days, please be patient. Thanks!

Always patient :nerd:   You’re busy, it’s totally understandable!

I have same problem here “

On NetApp storage arrays, all IntelliSnap backups of network shares use SnapDiff. You can also use SnapDiff for regular backups of network shares. 

SnapDiff is especially effective when you have a large number of files that do not change a lot between snapshots.

 

how do you enable snapshot on volume without IntelliSnap? 

@MaxJamakovic , confirming, is this the issue you have in the below thread?  I'm working on getting someont to reply:

 

Yes it is! Thanks Mike

Hi, just logged Incident 211123-170 for this. I think it should list all our issues.

Sharing the resolution:

- Removed the previously added Cluster info from the array management.
- Created a password for user on NetApp.
- Added the array using the backup lif IP and created a host alias for it in array management.
- Started a Full backup of the client and a snapshot was now created on the volume.

Just to give a kick to this topic, we noticed that SnapDiff V2 is working with the above configuration, but SnapDiff v1 and V2 will be discontinued by NetApp starting OnTap 9.10.1.

 

From that moment on you will have to use SnapDiff V3 which also brings some additional requirements being:

  • the info supplied in array management needs to be from a NetApp LIF which has the cluster management role
  • the user credentials in array management need to be from a user which has cluster admin rights to be able to create the required access tokens for SnapDiff V3

 

We have found this out from the Commvault logging were we found that SnapDiff was falling back from V3 to V2 during a back-up: Please verify that array yxxx.xxx.xxx.xxx] is a cluster interface and login nvsadmin] has a role on the file server which allows access_tokens.

 

And from a NetApp KB article which states that for SnapDiff to work it needs:

  • to target a LIF with a default-management policy
  • to use a valid snapdiff user on SVM

 

Would it be possible to have BOL update so these requirements are clear on forehand

 

@Luis Constantino , is there a specific document you’d like tos ee updated?  Each page has a little chat button in the upper right  to give feedback.  This creates a work item for our docs team to review and get your changes addressed.

Reply


Terms & ConditionsCookie settings