Based on the link you provided licenses are consumed in 500GB chunks so why not have 6 jobs of 500GB each for the 3000GB you have in the file shares?  I don’t back up NAS stuff much so take this as just a suggestion based on that link.

Hi Chris, thanks for the response.

I don’t think you would want to or even can backup partial file shares in a job. Even if you could, I am still curious as to how the licensing would be consumed if each job would sum to <500GB and then round down to 0 or if the licensing would be based on the total 3000GB within the filer no matter the breakdown and split of shares and jobs.

I did not mean back up partial shares just back up what is close to 500GB in each job but full shares.  It seems like things are in 500GB chunks for licensing based on your link as they give an example of a job with 1100GB which rounds down to 1000GB but multiplies by 2.  Not sure the answer here as I am also not a licensing expert so hopefully someone with more details can chime in.

I have an answer for you.

Source: https://helpcenter.veeam.com/docs/backup/vsphere/nas_licensing.html?ver=120
 

• If different data sources are protected with different file backup or object storage backup jobs or with different file backup to tape or object backup to tape jobs, Veeam Backup & Replication rounds the protected amount to 500 GB separately for each data source and calculates the number of instances required to protect each of them. After that, it sums up the total number of license instances to consume for unstructured data backup or for unstructured data backup to tape.

For example, if the amount of data is 1100 GB and it is protected by 2 backup jobs simultaneously, Veeam Backup & Replication will round it down to 1000 GB and multiply the amount of data by 2. As a result, a total of 4 license instances will be consumed.

The way I read it, it’ll round down to the nearest 500GB with a 500GB minimum.  Veeam is pretty generous, but I think the’d see through people gaming the system by backing up multiple sources less than 500GB to avoid using a license, so I’m fairly certain that first 500GB consumes one license, regardless of if it’s 300GB or 3GB.

• Veeam Backup & Replication rounds the protected amount of data for each file share or object storage down to the nearest 500 GB.
• One license instance covers 500 GB of the protected amount of data.

Assuming the above, a 300GB backup would consume 1 VUL.  If you have 10 300GB backup jobs running, it’ll consume 10 VUL’s.  See the below:

• If different data sources are protected with different file backup or object storage backup jobs or with different file backup to tape or object backup to tape jobs, Veeam Backup & Replication rounds the protected amount to 500 GB separately for each data source and calculates the number of instances required to protect each of them. After that, it sums up the total number of license instances to consume for unstructured data backup or for unstructured data backup to tape.

If you have the same job backing up all 10 300GB sources on the same root share, it’ll sum the data protected, and then round down to the nearest 500GB, and then calculate VUL’s.  In the this case, 6 VUL’s.  See the below:

• If an unstructured data backup job protects several file shares residing on the same NAS device (the same share root or NAS filer), the approach to calculating license consumption depends on how the file shares are added to the infrastructure:
  • If you have added the whole share root (\\root\) to the infrastructure and the file backup job protects shares \\root\share1 and \\root\share2, Veeam Backup & Replication first sums the protected amount of both file shares, rounds it down to 500 GB, and then calculates the number of license instances to consume for file backup support.

If you have the same job backing up all 10 300GB sources with each share specified individually, regardless of being at the same root, it’ll sum the data protected for each share separately, and then round down to the nearest 500GB, and then calculate VUL’s.  In the this case, 10 VUL’s.  See the below:

• If an unstructured data backup job protects several file shares residing on the same NAS device (the same share root or NAS filer), the approach to calculating license consumption depends on how the file shares are added to the infrastructure:
  • If you have separately added shares \\root\share1 and \\root\share2 to the infrastructure and the file backup job protects both of them, Veeam Backup & Replication first rounds the protected amount for each file share down to 500 GB, separately calculates the number of license instances to consume for each file share, and then sums the license instances to calculate the total number of license instances to consume for file backup support.

I also don’t do anything with NAS backups, but that’s the way I interpret the link that Chris posted above.

I agree with Derek, with exception of the rounding in the separate share & job configuration.

 

You have a single NAS device, but you have added 2 of its shared folders \\root\share1 (490 GB) and \\root\share2 (600 GB) as separate file shares to the inventory. The file shares \\root\share1 and \\root\share2 are added to one file backup job. In this case, Veeam Backup & Replication first rounds the protected amount for each file share down to 500 GB, separately calculates the number of license instances to consume for each file share, and then sums the license instances:

490 GB ~ 0 GB = 0 license instances

600 GB ~ 500 GB = 1 license instance

0 + 1 = 1 — protection of 2 file shares in this case consumes 1 license instance.

all other scenarios that I can think of right now will add the shares/job together and will divide that number by 500GB  = license and if its not a round number is will round down. 1490GB = 1000GB rounded  = 2 licenses.

Wow, thank you all for the great insight you are providing.

One thing to note, on that linked page, it specifically says data <500GB consumes 0 licenses:

If the protected amount of data for the file share is 499 GB or less, Veeam Backup & Replication rounds it down to 0 GB. In this case, protection of this file share will not consume license instances.

I guess the final question then is what is considered a “data source”. Because if each volume in a NetApp is considered a separate data source, then based on my understanding of the above, the data should not be summed before licenses are calculated and instead each round down to nearest 500GB then sum together then consume licenses which would be much less than if all data was summed first before rounding.

Scenario 1: There are 10 300GB volumes. Assume each volume is a separate data source.

• Each volume rounds down to 0GB
• 0 VULs are consumed

Scenario 2: There are 10 300GB volumes. Assume each volume is considered the same data source.

• Volumes sum together to 3000GB
• 6 VULs are consumed

I am hoping scenario 1 is how it works 

Wow….didn’t realize they wouldn’t charge a VUL for <500GB.  Good on Veeam though.  They’re pretty generous on things like that.  Same goes for rental licensing for Veeam Backup for M365.  Each mailbox consumes 1.5 points, and if you have a half point, such as 37.5 (25 mailboxes), it’ll round down to 37.

Scenario1 only works if you add the shares individually and create separate jobs. If you do 1 job with all separate shares its will count them all 10x300GB in 1 job = 3000GB = 6 licenses.

I mean….reading this, it sounds like it would consume nothing.

• If an unstructured data backup job protects several file shares residing on the same NAS device (the same share root or NAS filer), the approach to calculating license consumption depends on how the file shares are added to the infrastructure:
  • If you have separately added shares \\root\share1 and \\root\share2 to the infrastructure and the file backup job protects both of them, Veeam Backup & Replication first rounds the protected amount for each file share down to 500 GB, separately calculates the number of license instances to consume for each file share, and then sums the license instances to calculate the total number of license instances to consume for file backup support.

Assuming you set this up, please report back which one you find it to be consuming.  I would have had said 6, but if <500GB is 0 instances…..

The way I read it, if they’re in the same job, and on the same NAS root or filer, but each share is separately added, it will count VUL’s for each share and then add the VUL’s together.

Correct, but when you add the shares (datasources) 1 per job and you stay under 499GB its 0

 

Case 3

You have 2 file shares File Share 1 (990 GB) and File Share 2 (890 GB) each protected with a separate file backup job. In this case, Veeam Backup & Replication rounds the protected amount of each file share down to 500 GB, calculates the number of license instances required to protect each of the file shares. After that, it sums the calculated number of license instances required to protect the shares:

990 GB ~ 500 GB = 1 license instance

890 GB ~ 500 GB = 1 license instance

1 + 1 = 2 — protection of 2 file shares with separate file backup jobs consumes 2 license instances.

I think Derek is right here, I totally missed where it called out “NAS Filer” in “If an unstructured data backup job protects several file shares residing on the same NAS device (the same share root or NAS filer)” which is exactly my use case.

Thank you all for helping me comprehend this doc! I will report back once I get Veeam set up and confirm the behavior since we will now likely go through with it.

We would like to hear back what the results will be!

The question would be if Veeam will know that all the shares belong to the same filer.

Hmm...the way I’m reading it, it’s the same if you do one per job, or each share individually specified all in one job.  The only time I see it summing up all the shares and then calculating VUL’s is if you specify the root on a job that covers all shares.

• If an unstructured data backup job protects several file shares residing on the same NAS device (the same share root or NAS filer), the approach to calculating license consumption depends on how the file shares are added to the infrastructure:
  • If you have separately added shares \\root\share1 and \\root\share2 to the infrastructure and the file backup job protects both of them, Veeam Backup & Replication first rounds the protected amount for each file share down to 500 GB, separately calculates the number of license instances to consume for each file share, and then sums the license instances to calculate the total number of license instances to consume for file backup support.

• If you have separately added shares \\root\share1 and \\root\share2 to the infrastructure and the file backup job protects both of them, Veeam Backup & Replication first rounds the protected amount for each file share down to 500 GB, separately calculates the number of license instances to consume for each file share, and then sums the license instances to calculate the total number of license instances to consume for file backup support.

Exactly so
nas share1 300GB  = 0 license (round per share first)
nas share2 300GB = 0 license (round per share first)

Total it by summing up total license instance 0+0 etc.

Thats my take on it ;)

Hey @hunter.britton i did a write up for this over a year ago looking at all the file share backup options available within Veeam and how it impacts licensing, it explicitly discusses the license scenarios around storage utilisation and how Veeam calculates VUL consumption.

http://micoolpaul.com/2023/05/18/veeam-file-shares-licensing-feature-comparison/

Thanks Michael. Your wording there clarifies some of the other confusion I had with the doc. I guess I am just unsure as to if the NetApp volumes are considered separate distinct file shares or if they are considered under the same root since the NetApp is added via the Veeam NetApp integration using the API rather than manual NFS/CIFS paths.

Well I’ve got an answer. I am backing up 3 volumes from the NetApp filer, each sized at about 320 GB but only containing 300GB of data in the form of a sparse file. 

Backup job 1

• All 3 volumes added in the same job
• Consumes 1 VUL

Backup job 2

• Separate job created for each volume (3 jobs total)
• Consumes 2 VUL

This behavior seems off from what we’ve read and understood. I would expect both scenarios to consume 0 licenses.

What I expected based on my reading/interpretations.

Not what I expected….wonder why….

Great post...thanks for sharing this info.

Probably because you used the netapp plugin, is it possible you can just add the smb or nfs share by ip or fqdn to the inventory?

Edit typo

I’ll give this a shot and see if adding them manually has any different behavior.

One more update for you folks:

I am backing up 3 volumes from the NetApp filer, each sized at about 320 GB but only containing 300GB of data in the form of a sparse file. 

Backup Job 3:

• Each volume is added via an NFS share with the direct path (ex: netapp01:/share1) separately and added all to the same job
• Consumes 0 VUL

Backup Job 4: 

• The root NFS share is added to inventory (ex: netapp01:/) and all 3 sub shares are added to the same job
• Consumes 1 VUL

This all makes sense and seems to exactly follow the docs. 

Based on all I’ve tested, it seems like adding shares using the NetApp plugin is the same as adding a root share and backup up the sub-shares.

Thank you all for your input!

Hi Hunter,

Yes the plugin will see all those share and mark them as 1 root. 
A single share won’t / can’t check that.

Good to hear this worked out for you!