V10 NAS Backup License Cost

[selva]

The pricing of NAS backup is indeed baffling. If the NAS device supports iSCSI one could mount it on a Windows server and serve shares from it. That allows virtually unlimited amount of data backup with just one instance. Am I missing something? NAS backup is still useful to have if the price was more affordable.

A second reason for avoiding NAS backup in small setups would be the lack of support for NAS backup (or copy job) to rotated drives which is a very convenient medium for regularly moving backups offsite in small shops.

[Gostev]

If image-level backups work for you, then by all means you should continue using one for virtually unlimited file protection.

Our NAS backup addresses the very specific needs and requirements of enterprise-grade file-level backup, for example maintaining long-term file versions archive. This is something that is simply not possible to do with image-level backups in a cost-efficient manner, but then again - not everyone has these requirements to start with.

For the same reason, there's no rotated drives support in the initial release, as this offsite backup method is not commonly used by enterprise customers. However, if there's significant demand for our NAS backup from SMB clients, then we will certainly add support for rotated drives as well. It's not that hard.

[Gostev]

Hello, this is an interesting use case we have not thought about, so let me discuss this internally. Right now, we don't treat such servers any differently from regular NAS.

@DerOest in the end, we decided to make your use case of additional "safety backups" of those most important files completely free. This was the easiest way to approach it, as matching workloads and file shares is not always possible. For more information, see the What's New in 10a > KB3228 (see Licensing section). Thanks for your feedback!

[Gostev]

It's a fair feedback. We'll look at making NAS backup more affordable for Veeam Essentials customers specifically, considering they are small businesses.

@dasfliege based on the feedback from you and other users and partners, we decided to provide Veeam Essentials customers with the ability to buy NAS capacity at the new low price of USD 100 per 1TB. Minimum purchase is 5TB, maximum is 50TB. This option will be on the price list starting August 1st.

Note that this is a separate license counter, so it is not a part of the existing 50 VULs limit of Veeam Essentials. In other words, when purchasing NAS capacity this way, Veeam Essentials customers will be able to protect up to 50 machines and up to 50TB of NAS data at the same time.

[dasfliege]

That's actually great news. Thank you very much!

[sdv]

To recap this topic, maybe I misread something:
V10a doesn't change license costs for Enterprises who require long-term file versioning for their large file servers, as they are still paying 1 instance license per 250 gb for file share backups.

The thinking was:
2. Industry-average VM size is 150GB, among which 25GB is redundant/useless data (OS files). In other words, 1 VUL protects 125GB of actual useful production data (when data resides in a VM).

But as always, we will monitor feedback closely (and also sales results), and will adjust this ratio if necessary in the following product updates.

To scale these 250 Gb licenses based on average VM size doesn't make a lot of reasonable sense as file servers are always much larger than the Industry-average VM.
Therefor if one needs long-term versioning for their Enterprise file server(s), 250 Gb of data feels like an underestimated number.

It makes this feature the most expensive one to use.

[wishr]

Hi Stefan,

Thank you for the feedback.

We encourage all enterprise customers to discuss pricing aspects of the NAS offerings directly with our sales representatives. There are various discount options available for enterprises that should satisfy your requirements and at the same time fit the budget.

Thanks

[Gostev]

To recap this topic, maybe I misread something:
V10a doesn't change license costs for Enterprises who require long-term file versioning for their large file servers, as they are still paying 1 instance license per 250 gb for file share backups.

Mostly correct. v10a does introduce a minor change in that it does not charge for incomplete 250GB data sets any longer, which will result in a small decrease of license consumption. But other than that, there are no changes with VUL-based NAS licensing.

To scale these 250 Gb licenses based on average VM size doesn't make a lot of reasonable sense as file servers are always much larger than the Industry-average VM. Therefor if one needs long-term versioning for their Enterprise file server(s), 250 Gb of data feels like an underestimated number.

Yes, I do realize that file servers are always much larger than the industry-average VM. However, why does the capacity of a device storing the data matter in the first place? IMHO the only thing that really matters is the value of actual data to the business. This is what determines whether to back up the given data at all, and if yes - at what frequency and redundancy level.

Or, are you saying that the production data stored in file servers is somehow less valuable than the production data stored on VMs, and so the cost of backing up 250GB of data from file servers should be cheaper than the cost of backing up the same 250GB of data from a VM? This would be an arguable statement.

[sdv]

Yes, I do realize that file servers are always much larger than the industry-average VM. However, why does the capacity of a device storing the data matter in the first place? IMHO the only thing that really matters is the value of actual data to the business. This is what determines whether to back up the given data at all, and if yes - at what frequency and redundancy level.

If storage capacity doesn't matter, than it could've been 1 license per fileshare .

File servers usually contain a bunch of department folders where the actual value rate of the data is fluctuating and 'randomly' placed in and around folders. Its not determined by a specific share. To be safe, you want to backup a top level share. When time passes even important files are forgotten until one needs them. Most data is probably too old or not of any value in case of DR strucks.
Which means, arguably, alot more data is being less valueable than VM data.

Or, are you saying that the production data stored in file servers is somehow less valuable than the production data stored on VMs, and so the cost of backing up 250GB of data from file servers should be cheaper than the cost of backing up the same 250GB of data from a VM? This would be an arguable statement.

Yes. Most likely 80-90% of the files from fileservers are not business critical in case disaster strucks.

However it is also the place where most production restores are taking place due to user incidents, short and long term, high and low criticality. File level restores using versioning adds value for endusers and backup operators. But the price-value for this feature is too steep compared to regular VM backups.

[Gostev]

Most data is probably too old or not of any value in case of DR strucks.
Yes. Most likely 80-90% of the files from fileservers are not business critical in case disaster strucks.

It's not just about disasters though? Many customers (and almost all enterprise customers) are required to keep the history of file versions for reasons other than mere disaster recovery. And this requirement is hard to achieve with image-level backups from economical perspective.

Which means, arguably, alot more data is being less valueable than VM data.

I don't necessarily agree to this statement, as I know many customers who are the opposite: have their most business critical data sitting on NAS. So at the very least, this varies from customer to customer. And we definitely expect that some customers will choose not to backup their NAS at all, because the value of data does not justify the cost of backup and (even more so) the cost of secondary storage.

But in any case, it sounds like we're already there with the current pricing anyway? Because effective cost of backing up NAS data with Veeam is a few times cheaper than backing up VM data. While it's rarely a few times less important, at least in cases when you do actually need to back it up.

[sdv]

@Gostev, Thanks for replying so quickly!

It's not just about disasters though? Many customers (and almost all enterprise customers) are required to keep the history of file versions for reasons other than mere disaster recovery. And this requirement is hard to achieve with image-level backups from economical perspective.

Certainly not only disasters, also incidents, enduser errors. Which is what I said here;

it (it = fileserver restores) is also the place where most production restores are taking place due to user incidents, short and long term, high and low criticality. File level restores using versioning adds value for endusers and backup operators. But the price-value for this feature is too steep compared to regular VM backups.

E.g. 8 TB file servers with fileshare backup costs 36 licenses for this feature. Having multiple file servers (home drives, department drives etcetera) with this size and the demands for versioning doesn't make it better. Unless going with VM backups and copy jobs to build a longer retention, than its only 1 license per server.

Have their most business critical data sitting on NAS. So at the very least, this varies from customer to customer. And we definitely expect that some customers will choose not to backup their NAS at all, because the value of data does not justify the cost of backup and (even more so) the cost of secondary storage.

Many customers many requests, many demands . The file share backup feature is not only being used backing up dedicated NAS boxes . It is also used for fileservers running as a VM while its storage is also running on the same box, not a seperate NAS box. Why? Because versioning; Identical as some other tools out there in the wild.

Personally I'd rather see the licensing at 1 TB per license or X amount per fileshare to be size independent, maybe more future proof?

[Gostev]

Personally I'd rather see the licensing at 1 TB per license

Well, the good news is that the effective price is even cheap than "1TB per license" for Veeam Essentials now, as you can see from my post above. While larger customers who don't fit Veeam Essentials have always had access to significant volume discounts.

[username0815]

We are switched to SyncBack for ~56€ - it does the filecopy jobs cheaper and better then the beta construct of veeam file copy backup. And a wonder - I have not to pay a licence for 'every file'

[Gostev]

If it achieves your RPO and RTO requirements for the size of data you have, why not? I've seen plenty of our customers using even cheaper options

For example, we just sold 1PB of NAS capacity here in Switzerland to one well known manufacturer. They used robocopy to make dumps of their NAS for years, which is completely free! But it was too slow for their volume (incremental backup took days on such a volume of data, so they could only do monthly dumps pretty much). Meaning, no versioning was possible even remotely. But now with Veeam, incremental backup of the same share takes 15 minutes, and they can capture every file version that is created.

As I said before, our NAS backup is targeted at the Enterprise file backup needs and capacities. So, there's no point in comparing its cost with basic file sync solutions. If you have basic needs and a small NAS, then you can't beat free (our File Copy jobs). Then there are some cheap but more advanced options in the middle, like the one you shared - this can be a good solution for customers like yourself with little data to protect (as far as I remember, you are using something like QNAP of FreeNAS).

While we provide a high-end solution for a different market. So, everyone can choose what's best for them!

[Gostev]

Although I should probably mention, now that 10a allows to protect file servers and file shares under 250GB in size with NAS backup jobs for free, it will certainly shake things up on that very low end of the market!

[Samba222]

Feedback: here in southern germany is the purport of a lot of medium sized companies that Veeam does the most unattractive pricing with NAS Backup in compare to older, established enterprise solutions (with even more functionality, File-Server around 150TB)
...so yeah, actually no chance to establish this as a "business case" - although we are otherwise very pleased with Veeam - and we have to look for other solutions.

[Gostev]

Actually, there's the comparison of MSRP with those "established enterprise solutions" on the previous page, which shows that Veeam is a few times more attractive than those from price perspective. So, the only possible explanations to your feedback is that either those vendors discount very heavily, or use back-pocket licensing options. In either case, this becomes the volume discount negotiation issue (not something I can control).

[jmmarton]

IYes. Most likely 80-90% of the files from fileservers are not business critical in case disaster strucks.

This is where an enterprise storage management policy beyond just backup is important. There are various vendors who provide reporting tools into unstructured data. Once you have that type of insight, e.g. shares where most files haven't been modified in X days/weeks/months, then you can setup a backup policy that only protects business critical data. This keeps backup costs low regardless of whether you're using Veeam or anyone else.

Unfortunately most enterprises don't have that type of insight into their data. The "easy" answer then is to backup everything. And to Anton's point, this is where working with Veeam sales is important as there are various discounts that can be leveraged when talking about hundreds of terabytes or even petabytes of data.

Joe

[droorda]

One of the best parts of veeam is that in a vmware environment I do not have to worry about which VM's are "business critical" and which ones are just nice for the business to keep around. As long as I license the compute they all just role into a single job that is easy for me to monitor (and if need be restore from).
Given Veeam's "keep licensing simple and flexible" statement, would it not be easier to just allow us to use something like a vmware CPU license to fully license a single nas device(interface) of any size, just like we can license a vmware host of any size? This would create a practical performance limit on size based on how often I need backups taken, but allow those with static data or small budgets the ability to backup large datasets the same as they would in Vmware where a cpu is a single performance limiter.

[Gostev]

This approach won't work simply because NAS devices vary in capacity by a few orders of magnitudes, unlike VMware host CPUs. So we would be be the same both for 10TB QNAP/Synology, and for 1PB Isilon/NetApp. While support costs for these environments are dramatically different.

[droorda]

Thanks for the quick response. I understand you perspective, being fair to small and large users and getting a return on your developer time is important.
From my perspective, My Esx host storage can also range from a few TB to a few PB while still maintaining a consistent Veeam cost. The only issue on my side then is time to take backup and how many CPUs the servers have.
Having an Isilon they do have a node size limit, billing based on node count would be a similar limit/cost to what happens in ESX. I do understand that it is not easy to look into a storage unit to get its hardware layout, which is likely part of why a server license is not possible for NAS units.
Another thought on licensing would be to bill based on changing data. I have a significant amount of data(100TB) that is years old(static) that I would like to keep in Veeam instead of an uglier/other backup product that has no size limits. I don't need the same level of protection of this data, but do still need to protect it. I am testing the File to tape option, but do not think it will have the needed performance to finish before I retire. Not surprising since this is not what it was made to do. The other issue with this is requiring that the destination is a Tape or VTL.
While I know any changes to the license model would be difficult to do without making current users angry. I am just putting these out as ideas that would make this product an option for more of us fans of Veeam who have budget limitation.

[Gostev]

I think there's no need to change the licensing, as basing it on the amount of protected data is really the only fair way. We just need to tweak the effective price according to what most customers are willing to pay, based on our observations of discount levels this year. In other words, give more GB per VUL. With this approach, no one should be angry hopefully.

[selva]

We have two NAS shares backed up, one about 350GB and another < 1GB. After updgrading to 10a I expected the NAS instance license consumption to go down to 1, but it uses up 2 instances (count = 2, multiplier = 1, instances = 2). The manage window shows the two shares as 0.1 and 1.4 instances. Shouldn't that add up to 1, as fractional 250GBs are no longer counted? Or, am I reading the new license rules wrong?

[Gostev]

Unused licenses should be released after 30 days as per standard VUL logic: they are not released immediately after the protected workload is gone, because it may have been deleted accidentally and will need to be restored a few days/weeks later, thus effectively realizing the value from our product. To account for legitimate deletions, we allow to exceed the license by 10 instances or 10% of total instances (whichever is greater).

[dandav]

Something that would be good for Veeam to consider is a separate TB allowance in the VUL for direct-to-archive type backups for files that never change or some sort of forever incremental license by the amount of changed data backed up. For example, we have a lot of video files from cameras which are recorded once and never change, we back up new files on creation to an immutable backup destination (LTO WORM tapes) in what is essentially a forever incremental backup job.

We are currently looking into migrating to Veeam and the VUL NAS pricing is fine for our standard business documents on NAS but as it stands we will still need to run separate software for backing up media files as the 250GB VUL limit makes Veeam noncompetitive for this scenario. I know we can kick and scream for some deep discount level but that doesn't stand up in our business. We've been bitten before when a vendor provides deep discounts on initial purchase then refuse to offer the discounts when time comes to expand. With any capacity based licensing purchase I need to provide potential future expansion scenarios built around MSRP in the proposal or it won't even be considered.

[wishr]

Hi Daniel,

In fact, we do have such a per-TB offer for very large volumes of data, and it has a few times cheaper MSRP comparing to VUL. Please, contact your Veeam sales representatives directly for more information.

Thanks

[Jonathan]

As a follow-up: I just received an updated quote to back up our 900TB NAS appliance. It boiled down - hang on to your seats - over € 200.000. Just for the first year. Before you ask, this was after nagging our Veeam rep for about a year.

The entire 8-node DellEMC Isilon cluster we need to back up didn't even cost that much. Any reseller or Veeam employee who tries to justify this kind of pricing strategy needs to get their head examined. I'm sure it's a quality piece of software, but the value to the customer is maybe 15-20k. Tops.

[Gostev]

Justifying the pricing strategy is not the same as justifying the specific quote though? I wouldn't justify what you've been quoted with. Because for your amount of data, there is the special capacity-based SKUs available that covers 1PB of NAS, and this SKU is more than twice as cheap as what you were quoted with (and that's MSRP, so before any discounts).

[Jonathan]

In that case Anton, your colleagues at the Veeam Benelux really need a wake-up call. I've been asking them for over a year now about these licenses, so there's no excuse.

[Gostev]

Thanks, I will pass the feedback to their management.