NAS backups and VUL consumption

[mark49808]

I'm trying to get clarity around how VULs are actually consumed. I think I can best express in an example with two cases how i Imagine this would work. Is it case 1 or 2?

Example scenario:
NetApp share is 1TB. Data churn happens where share content changes 10% every day. New unique data is added and old data removed. Think of a daily log dump that has older data removed on a daily basis, while Veeeam is keeping the old data protected for a long duration (years maybe).

Case1: 2 VULs are consumed max as the share itself is only 1TB and that is the determining factor.
Case2: 2 VULs are consumed initially, but then over time that adds up to a significant VUL consumption as the unique data is "additive" and VULs are consumed based on "protected" data. 1TB + 10% + 10% + 10% + 10%... and so on.

I'd hope for case 1.

[jorgedlcruz]

Hello Mark,
I am afraid it will be Case 2, look you can find this very same example on the next link, on this case under Case number 4.

You can always negotiate this topic with your Veeam Sales representative, and if I recall correctly, there are some NAS Only SKU for really big Datasets, like 500TB, 1PB, etc. So you can calculate ahead of your year's retention.

Veeam will do some compression, and depending on where you store the data, that destination will do some deduplication as well to keep things small.

But regarding Veeam licensing, it is case 2.

Let us know! Hope it helps.

[mark49808]

Now i'm confused more.

From the article you linked:

"Two days later, the size of the file share increases to 1510 GB. Veeam Backup & Replication runs the file share backup job and recalculates the number of license instances to consume based on the increased size of the NAS share:

1510 GB ~ 1500 GB = 3 license instances"


What I'm saying in my case is the "size of the file share" NEVER increases. Its always 1TB. The data within will be deleted, and new data will be added, but the actual share as presented to the world is 1TB. Per the language in this case it seems to revolve round the "size of the NAS share" which is not changing in my case.

Semantics? or Is there a difference?

[jorgedlcruz]

Interesting, I might be wrong, let me double check to confirm what I am saying is true, it is indeed a particular use case.

I will get back to you soon.

[mark49808]

Thanks. i get its an interesting use case. I consider it more of an "Archive" than a backup. But still the use remains. Sometimes you need files to exist only on secondary storage for compliance reasons so that results in a high churn for your file share with a bunch of unique data.

[jorgedlcruz]

I know that we readjust data protected every 30 days for NAS after deletion, you can force it every day with a powershell script https://helpcenter.veeam.com/docs/backu ... rvers.html

So perhaps that will be your answer. Need to verify it, as we might be already checking this specific use case.

Will come back with an answer, and test, tomorrow.

[Dima P.]

mark49808,

File share size stays the same despite the fact that some files were added/deleted between job runs. If your file share is 500GBs during first job run, then you delete 300GBs, then add 300GBs and start backup job the licensed size will be still 500GBs. Hope it helps!

[mark49808]

One additional question here. If I add a File Share backup per this screen, which of these options consumes NAS Backup capacity licenses?

https://helpcenter.veeam.com/docs/backu ... _share.png

I'm guessing the first one (File Server) would potentially not given it installs an agent on the server, and would be counted a simply a single instance? Or is that mistaken?

[Mildur]

I'm guessing the first one (File Server) would potentially not given it installs an agent on the server, and would be counted a simply a single instance? Or is that mistaken?

As long a "File Server" is backed up by a File Share Backup Job, NAS Capacity licenses are consumed. 500GB / 1 instance.
You can use a VM Backup Job or a Veeam Agent Backup Job for that "File Server", then it will only consume 1 instance for the entire file server.

Thanks
Fabian

[mark49808]

Thanks. What's the down side to using an agent-based license then? I cant seem to find a chart that shows what i'm missing/gaining by using NAS backup vs Agent. For Windows servers or linux servers, it seems there are a number of limitations that make the Agent-based better, mainly around it writing to a SOBR and taking advantage of the immutability and tiering avaialbie (s3-> glacier, etc). Where NAS has OTHER benefits such as instant share restore, but cant be immutable, and cant go to glacier but has a per 500GB instance cost. Has anyone published anything showing which to use when and why or at least describing all the tradeoffs?

[Dima P.]

mark49808,

I'd say agent backup or VM backup is the recommended way to process a workload whenever you have access to the hypervisor or guest operating system: backup the operating system level is faster due to changed block tracking and provides incremental backup at a block level (you do not process entire file after it's modification). Nas backup engine covers the cases when you do not have access to the hypervisor / guest operating system for image-level backup or you need to process a file share with unstructured data.