Netbackup VmWare restore change VMDK hash

[rfm_columbia]

I saved the .VMDK hash before execute a VmWare backup, but when I restore the same VmWare the .VMDK hash is another.

Is this normal?

Why Netbackup change the VMDK hash?

Could someone explain to me how Netbackup performs the recovery of a VmWare machine?

Perhaps there is some process in the VmWare restore flow that causes this change.

[X2]

Was the restore of the VM successful? i.e. it ended with status 0?

Did you automatically start the VM after restore?

[Michal_Mikulik1]

Hello,

NetBackup calls VMware to initiate snapshot creation of the backed up VM, so maybe this step writes some metadata into VMDK header. However to confirm this is a question for some VMware guru rather than for NetBackup forum.

regards

M.

[rfm_columbia]

The restore finish with status 0, successful!

We not automatically start the VM after restore, the VM still power off, but the hash is diferent.

[rfm_columbia]

We have two hashes, one hash before the backup and the other after the restore, but even when we do multiple restores with diferents dates and restore features, the hash of the restore is always the same, but diferent of hash before a backup.

[X2]

---

@rfm_columbia wrote:

..... but even when we do multiple restores with diferents dates and restore features, the hash of the restore is always the same, but diferent of hash before a backup.

---

So, with the same has for multiple restores (from same backup image), it indicates that the restored vmdk is same in all cases (restores). It maybe different from the has of backup image file due to what @Michal_Mikulik1 was alluding to. I suggest to open a case and ask for an explanation if you want to be certain.

Otherwise, as far as restored VM and data is concerned, you can always verify by testing the presence of your data files and their corresponding checksums by performing restores.

[meuhassan]

Yes, it is normal for the VMDK hash to change after a NetBackup backup. This is because NetBackup uses a process called deduplication to compress the backup data. Deduplication works by identifying duplicate blocks of data and storing them only once. This can significantly reduce the size of the backup data, but it also means that the hash of the VMDK file will change after the backup.

When NetBackup restores a VMDK file, it first checks the hash of the file to see if it has been deduplicated. If the file has been deduplicated, NetBackup will restore the deduplicated blocks of data from the backup media. This means that the hash of the VMDK file will be different after the restore.

Here is a brief overview of how NetBackup performs the recovery of a VMware machine:

1. NetBackup identifies the VMDK files that need to be restored.
2. NetBackup checks the hash of each VMDK file to see if it has been deduplicated.
3. If the VMDK file has been deduplicated, NetBackup restores the deduplicated blocks of data from the backup media.
4. If the VMDK file has not been deduplicated, NetBackup restores the entire file from the backup media.
5. NetBackup mounts the VMDK files on a VMware datastore.
6. NetBackup creates a new virtual machine from the mounted VMDK files.

The process of restoring a VMware machine from a NetBackup backup is relatively straightforward. However, it is important to understand that the hash of the VMDK files may change after the restore, due to the deduplication process.

I hope this explanation is helpful. Let me know if you have any other questions.

[rfm_columbia]

Thanks for the explanation it was very enlightening for me.

[Michal_Mikulik1]

Hello guys,

you are confusing two different hashes. While @rfm_columbia is talking about client side data hash (something like get-filehash on VMDK I suppose), @meuhassan is mentioning deduplication hashing used on backup server side. One is file-level, another one is block-level, both have different vendors/purposes etc.

Check my post above, I suppose it is a clue. @rfm_columbia  try to calculate VMDK hashes for these different cases: VM is without snapshot, VM with a snapshot, and VM is after Revert from snapshot which is comparable to restore from NetBackup. 

Regards

Michal

[rfm_columbia · Accepted solution]

Hello guys I discovery why the VMDK hash changes.
If you do the backup with VMware Optimizations (Enable file recovery from VM backup, Enable block-level incremental backup, Exclude deleted blocks, Exclude swap and paging files)

If the Optimizations are unchecked before the backup the VMDK hash remains the same after restore!