NetApp Cloud Volumes ONTAP Reviews

It is managing services in our production environment that are in Azure. It provides file shares, both NFS and CIFS, that are used by other applications that are also in Azure.

NetApp Cloud Volumes ONTAP is part of the production environment of our company so the entire company, over 5,000 employees globally, is touching it somehow. It's a part of an application that has data that resides on it and they may consume that application.

Cloud Volumes ONTAP is great because of the storage efficiencies that it provides. When you look at the cost of running Azure native storage versus the cost of Cloud Volumes ONTAP, you end up saving money with Cloud Volumes ONTAP. That's a big win because cost is a huge factor when putting workloads in the cloud. We had a cost estimate survey done, a comparison between the two, and I believe that Cloud Volumes ONTAP saves us close to 30 percent compared to the Azure native costs.

Azure pricing is done in a type of a tier. Once you exceed a certain amount of storage, your cost goes down. So the more data you store, the more you're going to end up saving.

The storage efficiencies from the NetApp platform allow you to do inline deduplication and compaction of data. All of this adds up to using less of the disk in Azure, which adds up to savings.

We have two nodes of the NetApp in Azure, which means we have some fault tolerance. That is helpful because Azure just updates stuff when they want to and you're not always able to stop them or schedule it at a later time. Having two CVO nodes is helpful to keep the business up when Azure is doing their maintenance.

The solution provides unified storage no matter what kind of data you have. We were already using the NetApp platform on our on-premise environments, so it's something we're already familiar with in terms of how to manage permissions on different types of volumes, whether it's an NFS export or a CIFS share. We're able to utilize iSCSI data stores if we need to attach a volume directly to a VM. It allows us to continue to do what we're already familiar with in the NetApp environment. Now we can do them in Azure as well.

It enables us to manage our native cloud storage better than if we used the management options provided by the native cloud service. With CVO, all of your data shares and volumes are on the one NetApp platform. Whether you are adjusting share permissions on an NFS export or a CIFS share, you can do it all from within the NetApp management interface. That's much easier than the Azure native, where you may have to go to two or three different screens to do the same stuff.

The storage efficiencies are something that you don't get on native.

Also, because of the NetApp product, we're able to use the SnapMirror function and SnapMirror data from our on-prem environment into Azure. That is super-helpful. SnapMirror allows you to take data that exists on one NetApp, on a physical NetApp storage platform, and copy it over to another NetApp storage platform. It's a solid, proven technology, so we don't worry about whether data is getting lost or corrupted during the SnapMirror. We are also able to throttle back the speed of the SnapMirror to help our network team that is paying for a data circuit. We're still able to copy data into Azure, but we can manage the transfer cost because we can throttle back the SnapMirror. It's just very solid and reliable. It works.

And all of us IT nerds are already familiar with the NetApp platform so there was not a major learning curve to start using it in Azure.

NetApp also has something called Active IQ Unified Manager, and it gives us performance monitoring of the CVO from an external source. There are several people on my team that utilize the CVO and we each have a personal preference for how we look at data. The Active IQ Unified Manager is a product you can get from NetApp because, once you license your CVO, you are entitled to other tools. CVO does have resource performance monitoring built in, but we primarily utilize the Active IQ Unified Manager.

Beyond that, it provides all the great stuff that the NetApp platform can do, but it's just in the cloud.

I think this is more of a limitation of how it operates in Azure, but the solution is affected by this limitation. There's something about how the different availability zones, the different regions, operate in Azure. It's very difficult to set up complete fault tolerance using multiple CVO nodes and have one node in one region and one node in another region. This is not something that I have dug into myself. I am hearing about this from other IT nerds.

We've been using NetApp Cloud Volumes ONTAP for two years.

We had issues with Azure when they did maintenance on the nodes. They just do their maintenance and it's up to us, the customer, to make sure that our applications are up and data is flowing. When Azure does their maintenance, they do maintenance on one node at a time. With the two nodes of the CVO, it can automatically fail over from one node to the node that is staying up. And when the first node comes back online, it will fail back to the first node. We have had issues with everything failing back 100 percent correctly.

We have had tickets open with NetApp to have them look into it and try and resolve it. They've made improvements in some ways, but it's still not 100 percent automated for everything to return back. That's an ongoing thing we have to keep an eye on.

It is definitely scalable. You can add more disk to grow your capacity and you have the ability to add more nodes. There's a limit to how many nodes you can add, but you can definitely scale up.

Tech support is good. A lot of it depends on the technician that you get, but if you're not happy with one technician, you can request that it be escalated or you can request that it just be handled by another technician. They're very eager to help and resolve issues.

We had some issues with permissions and with getting the networking correct. But we had a lot of support from NetApp as well as from Azure. As a result, I would not say the setup was straightforward, but we got the help and the support we needed and you can't ask for more than that.

I've always found NetApp support to be accurate and good with their communications. Rolling out this product in Azure, and working with the IT nerds in our company and with Azure nerds, occasionally it does add another layer of who has to be communicated with and who has to do stuff. But my experience with NetApp is that they are responsive and very determined to get situations resolved.

It took us about a week to get everything ironed out and get both nodes functional.

We had done a PoC with a smaller instance of the CVO and the PoC was pretty straightforward. Once we rolled out the production CVO that has two nodes, that's when it was more complicated. We had a plan for getting it deployed and to decide at what point we would say, "Okay, now it's ready for prime time. Now it's ready to be put into production."

For admin of the solution we have less than 10 people, and they're all storage administrator analysts like me.

Our licensing is based on a yearly subscription. That is an additional cost, but because of the storage efficiencies that the NetApp gives, even with the additional cost of the NetApp license, you still end up saving money versus straight Azure native for storage. It's definitely worth it.

Make sure that you can stay operational when Azure is doing their maintenance. Make sure you fully understand how the failover and the give-back process works, so that you can deal with your maintenance.

NetApp Cloud Volumes ONTAP is where we host our NAS storage on which we keep our files, et cetera. We have three clusters of CVO, each serving close to 300 terabytes of data. We have our SQL backup workloads and the application data residing in it. We are using the tiering policy, which pushes the inactive data down to cold storage to help save on costs.

Cloud Volumes ONTAP is all cloud-based and we have our workloads on Azure.

At one point we were paying close to $80,000 a month for cloud resources, and now it's down to $25,000 to $30,000 after using the tiering.

Also, using Unified Manager we are able to resolve issues before they have an impact. For example, there were conditions where bulk operations were happening against a particular volume, and our business was also writing the data. We caught it using Unified Manager. The IOPS were low and there was a high latency, close to 1,500 milliseconds. We had a look at exactly what operations were happening and, before the user even reported it, we reached out to the team that was doing the bulk operations to stop whatever process they were running. That's just one example. We have had a lot of occasions where the tool has been really handy when it comes to proactive monitoring. 

And it's not only for proactive monitoring. The same tool is also used for a lot of root cause analysis.

The storage tiering is definitely the most valuable feature. With the pay-as-you-go plan, we can choose between standard and premium storage, but we use only premium for high performance. High IOPS and low latencies are the main features of the premium storage. With respect to tiering, the inactive data is pushed to a lower tier where the storage cost is cheap, but the access cost is high.

NetApp also has something called SnapMirror replications and that's how we replicate our data from production to the DR site, for our BCP. It has pretty solid solutioning for the replications so the SnapMirrors are pretty handy when it comes to BCPs.

In terms of cloud resource monitoring, we use Unified Manager and it's pretty cool. It has both Excel-based metrics as well as graphical representations, which give us a clear idea of which particular file systems have performance problems. We can go over the statistical information and it comes in very handy. At the same time, it has an alerting mechanism where any sort of conditions can be configured and alerts are then sent to your mailbox or your mobile SMS.

It definitely needs improvement with respect to clustering and with respect to more collaborative integrations with Azure. Right now, we have very limited functionalities with Azure, except for storage. If CVO could be integrated with Azure that would help. When there is any sort of maintenance happening in the cloud, it disrupts the service in Cloud Volumes ONTAP. 

If those could be rectified, that would be really good news because it would reduce the administrative overhead my team and I are facing.

I have been using NetApp Cloud Volumes ONTAP for close to two years now.

It is stable, at least with respect to Azure, which is what we use.

NetApp is scalable. When we initially started with Cloud Volumes ONTAP, it had a hard limit of 378 terabytes as its maximum capacity per cluster. Beyond that you couldn't expand. So we had to spin up another cluster. When that was almost full we had to get a third cluster. But I believe that in the recent build of CVO they have introduced the ability to stack one license on top of another cluster, so you can have infinite data per cluster. So there were challenges, particularly with vertical scalability before, but that has now been fixed in the recent release.

In terms of increasing our usage in the future, we definitely will if required. It gives us the flexibility to perform automations and it has its own encryption tools. Right now, we are using it for one particular region in Europe, but we do have plans to get it out to other regions as well, but that's not going to happen immediately.

We have used their technical support on a lot of occasions and they have been pretty helpful. I'm completely satisfied with the resolutions they have come up with. We have created more than a hundred tickets in the last two years. Those were all submitted initially, when we had an older version of CVO, but now we hardly create tickets with support because our team has the ability to administer it.

We had ONTAP before CVO. We also used Dell Isilon and SoftNAS before we migrated to CVO. We switched because we found that SoftNAS was not stable enough to handle the workloads. We often had problems with the applications.

The initial setup of Cloud Volumes ONTAP is pretty straightforward. We didn't have any sort of difficulties getting it spun up. It was also pretty quick. We had everything needed in Azure. It hardly took us three to four hours to have the entire environment set up and ready.

We did some architectural planning for setting this up and we got all the approvals and licenses well in advance, before we actually configured it.

When it comes to maintenance, it depends on what kind of coverage an organization wants. We are a team of four who administer NetApp clusters alongside the cloud resources. We have roughly 2,000 users.

We worked only with NetApp. We had one architect and one contractor from Professional Services.

Now that we have started using tiering, we could still actually save more costs, but we haven't gone to that specific area. We know it is definitely going to affect the performance if we keep all the data in the cloud tier. That's why we haven't. But Cloud Volumes ONTAP has the flexibility to dump all the data to the cloud tier to save every penny possible.

We have an annual license renewal for all the clusters. The license comes with annual training and with some Professional Services time. We have used all of those. I'm not sure if that's standard or it's an agreement between our organization and NetApp, but that's what we get as a part of our licensing.

The money you can save with CVO depends on what type of configuration an organization needs. They can also push all the data down to the cold tier. The pricing model for the Azure Cool Blob offering is pretty low compared to the premium or the standard. The cost of cold storage would probably be 10 cents per GB.

Very few of the cloud service solution providers have that tiering option. Tiering results in a lot of savings.

NetApp, on the whole, is a whole different tool for me. Two years back, when I started, when I had my hands on it for the first time, I found it pretty interesting. I would note its simplicity. It's simple and, at the same, time very powerful and able to handle any sort of storage workloads.

NetApp is really cool. If your organization is looking for cost savings, NetApp is the way to go.

Overall, I would rate CVO a nine out of 10. We had a lot of problems with NetApp, but those were in the very early stages. And NetApp always promises to upgrade their products and they actually listen to the customer's problems. We have raised a couple of defects with NetApp, and they have always been supportive, getting these resolved as soon as possible. 

The NetApp organization, on the whole, is pretty good. They're coming up with go-to-market products like Azure NetApp Files, etc., which is actually the beta version of Azure Sites.

The primary use case is to move age old data to the cloud.

It is deployed on the cloud.

The tool saves us time and money. Now, it's easy to retrieve data back, where you can go back and look at the statistics to study them. Because my company is focused on healthcare, there's no time limit on the retention of information. It's infinite. So, instead of having all our data on tapes and things, which takes many hours to try to retrieve information back. This is a good solution.

The migration is seamless. Basically, we shouldn't be spending a whole lot budget-wise. We would like to have something reasonable. What's happening right now is when we try to develop a cloud solution, we don't see the fine print. Then, at the end of the day, we are getting a long bill that says, "Okay, this is that, that is what." So, we don't want those unanticipated costs.

We use the solution’s inline encryption using SnapMirror. We did get Geoaudits and things like that. In other words, everything put together is a security. It's not like just storage talking to the cloud, it's everything else too: network, PCs, clients, etc. It's a cumulative effort to secure. That's where we are trying to make sure there are no vulnerabilities. Any vulnerabilities are addressed right away and fixed.

The solution’s Snapshot copies and thin clones in terms of operational recovery are good. Snapshot copies are pretty much the write-in time data backups. Obviously, critical data is snapshotted more frequently, and even clients and end users find it easier to restore whatever they need if it's file-based, statical, etc. 

The solution’s Snapshot copies and thin clones have affected our application development speed positively. They have affected us in a very positive way. From Snapshots, copies, clones, and things, they were able to develop applications, doing pretty much in-house development. They were able to roll it out first in the test environment of the R&D department. The R&D department uses it a lot. It's easy for them because they can simulate production issues while they are still in production. So, they love it. We create and clone for them all the time.

The solution helped reduced our company's data footprint in the cloud. They're reducing it by two petabytes of data in the cloud. All of the tape data, they are now writing to the cloud. It's like we have almost reached the capacity that we bought even before we knew we were going to reach it. So it's good. It reduces labor, because with less tapes, you don't have to go around buying tapes and maintaining those tapes, then sending them offsite, etc. All that has been eliminated.

Right now, we're using StorageGRID. Obviously, it is a challenge. Anything that you're writing to the cloud or when you get things from the cloud, it is a challenge. When we implemented StorageGRID, like nodes and things like that, we implemented it on our bare-metal. So the issue is that they're trying to implement features, like erasure coding and things like that, and it is a huge challenge. It's still a challenge because we have a fine node bare-metal Docker implementation, so if you lose a node for some reason, then it's like it stops to read from it or write to it. This is because of limitations within the infrastructure and within ONTAP.

How it handles erasure coding. I feel it the improvement should be there. Basically, it should be seamless. You don't want to have an underlying hardware issue or something, then suddenly there's no reads or writes. Luckily, it's at a replication site, so our main production site is still working and writing to it. But, the replication site has stopped right now while we try to bring that node back. Since we implemented in bare-metal, not in appliance, we had to go back to the original vendor. They didn't send it in time, and we had a hardware memory issue. Then, we had a hard disk issue, which brought the node down physically. 

It needs better reporting. Right now, we had to put everything one to the other just to figure out what could be the issue. We get a random error saying, "This is an error," and we have to literally dig into it, look to people, lock files, look through our loads, and look through the Docker lock files, then verify, "Okay, this is the issue." We just want it to be better in alerting and error handling reports. Once you get an error, you don't want to sit trying to figure out what that error means in the first two hours. It should be fixable right away. Then, right away you are trying to work on it, trying to get it done. That's where we see the drawbacks. Overall, the product is good and serves a purpose, but as an administrator and architect, nothing is perfect.

There's always room for improvement. Overall, it's still stable.

60 percent of our tape data is sitting in the cloud now.

There's a limitation to scalability. Right now, when you want to expand the initial architecture, we have to add additional loads just so it can handle the data without hurting the performance. Then, we have to go back and request for more licensing. It adds to our licensing, thus adding to the cost. In regards to scalability, unless you have a five to six year plan ahead, we can't say, "Great, we have run out of space. Okay, let's try to increase space." It's not like increasing volume.

Unless a much more experienced person comes, I think the print and tech guy is only reading what he sees on the website. He pulls up their code or whatever, because what we see when we open a case is already there is an automatic case that's opened. We see typical questionnaires, but nothing pertaining to the case. For example, you run out of space or high nodes, the technical support is sitting there asking us something else. Nothing to do with high nodes and the volume being down or offline. It's not relevant. It is a generalized thing. You have to sit down and explain to them, "This has nothing to do with the questions you're asking. It's out of context, so you might want to look again and get back with the proper input." That's a pain.

However, the minute we say, "It's very critical," we see a good, solid SME on the line who is helping us.

I'm not experienced as many of my colleagues. They're really frustrated. We did convey this concern to our account person and have seen a lot of change.

The company has always been a NetApp shop even before I entered the company. We continue to use it because of the good products. We do market research, obviously. We do see good products, and every year there is improvement. When we want to do hardware upgrades, it's still very good. The way we are trying to develop, it's very seamless for us and not a pain. 

We have never felt, "We are done with NetApp. Let's move onto something else." I love to introduce other vendors into the mix, just so it's not a monopoly. We still love NetApp as our primary.

It is a little complex. It's completely different from the regular standard ONTAP, and how you manage and the learning code. Half the time you get confused and try to compare it with a standard cloud. You start to say, "Oh, this feature was here. How come it's not there? That was very good there. How come it's not here?"

We used NetApp Cloud Manager to get up and running with Cloud Volumes ONTAP. The configuration wizards and its ability to automate the process was good. We liked it. It's all in one place, so you don't have to go around trying to use multiple tools just to get things worked out. You see what you have on the other side plus what do you have on your end, and you're able to access it.

Mostly, we did it ourselves. When we went to MetroCluster, we used their Professional Services. For the rest of ONTAP, we deployed it ourselves. It is pretty much self-explanatory and has good training.

Cloud is cloud. It's still expensive. Any good solution comes with a price tag. That's where we are looking to see how well we can manage our data in the cloud by trying to optimize the costs.

I do know our licensing cost to some extent, but not fully. E.g., I don't know overall how much we have gone over the budget or where did we put costs down just to maintain licensing on it. That part of it, I don't know. 

I know the licensing is a bit on the high-end. That's when we had to downsize our MetroCluster disks and just migrate to disks that were half used. We migrated into those just to reduce maintenance costs.

We use Caringo. It's object storage migration for age old data. It is a cheap solution for us, so that's why we use that. When we compared prices, Caringo was much cheaper.

Once we migrated everything to Caringo, there were challenges because it's another vendor, and then you're working with two different vendors. We started having issues, so now we use StorageGRID.

We chose NetApp because we already had the infrastructure. Adding additional resources and features into the mix is much easier because it's one vendor, and they understand the product. If we needed to add something and improve on the solution, it's much easier.

I would recommend NetApp any day, at any time, because there's so much hard work in it. It's more open and transparent. Nobody is coming from NetApp, saying, "We're going to sell this gimmick." Then, you view all the good stuff but begin to realize, "This is not what they promised." For this reason, I would recommend NetApp.

They make sure the solution fits our needs. It's not, "Okay, we'll go to the customer site and tell what we feel like regarding their products." Even if it fits or not, it doesn't affect that they've gone through the door. A lot of people do that. NetApp makes an assessment, then they make sure, "Okay, it does fit in."

The product: I would give it an eight (out of 10). The company: It's a six (out of 10).

We have not yet implemented the solution to move data between hyperscalers and our on-premises environment. It's just from our NetApps to the cloud, not from the hybrid. The RVM team is planning on that. So, they can have the whole untouched thing put on the cloud rather than being hosted on our data stores.

We use it mostly for distributed files. For example, one of our customers is in construction. So they have centralized NetApp storage and set up replication with the Cloud volume ONTAP. Several branch users access the Cloud instance. And whatever work that they do on the Civeo instant gets replicated to the client's data center on the on-premise NetApp storage. And they use GFC with Seavio for a seamless experience.

So a lot of these licenses are at the rate that is required for capacity. So they're they're able to reduce the license consumption and also the consumption of the underlying cloud storage. 

So one of the clients had an on-premise 100 terabytes. When that same data went to CVO, he was able to reduce it to somewhere around 20/25 dB. So he was able to reduce using DTO compression, the consumption of this underlying cloud storage. And, obviously, with the licenses, now you need only 25 terabytes instead of 100 terabytes.

There's not much scope for improvement. I think the solution is more restricted with the underlying cloud. The performance of the single instances depends on the performance of the underlying cloud resource. 

I have been using NetApp Cloud Volumes ONTAP for two years. 

NetApp is stable. 

NetApp is scalable. 

The technical support team is wonderful.

Positive

It is easy because Blue XP makes it very easy. Just a few clicks. It's a little bit tricky for customers who are very new to Cloud. While they are the traditional customers who use their data center, if they are not well versed with cloud, they might face challenges in setting it up because that's more dependent on public cloud vendor, not the NetApp.Sometimes the internal support renewal is an issue because the NetApp system sees it as an expired license. So when you try to renew it, it is not reflected in the NetApp portal. So the flow of the support could be improved.

For enterprise customers, it's a very cost effective. But in the SMB segment, yeah, pricing is a little bit challenge for your time.

I rate the overall solution a ten out of ten. 

I work in healthcare, so we store a lot of data on-prem and in the cloud. We are using multiple vendors and platforms. ONTAP was an excellent choice because everything is integrated. We haven't gotten into the AI features, but the security is pretty good. 

NetApp's Integration with AWS has helped us because we had a tough time transferring data when we used an ONTAP competitor as our storage partner. They don't have integration with AWS tools, so we had to figure it out on our own. 

ONTAP has built-in integration and allows us to replicate a copy to our second data center. Everything is in one channel. It's possible without the technology, but it's more time-consuming. NetApp saves us a lot of time, and the administration is simpler.

NetApp's support could improve.

We had Dell EMC. Dell EMC was good initially, but their customer support isn't so great.

I rate NetApp 10 out of 10.

We rely on NetApp Coud Volumes ONTAP for a wide range of purposes, including VMware, SQL, Oracle, and file storage. It serves as our go-to storage solution for almost every use case.

The transition to the AFF storage solution significantly improved our organization by reducing our physical footprint. We went from a FAS system with two controllers and twenty drive shelves to just four controllers and two drive shelves with the AFF. This led to substantial reductions in power consumption and space requirements in our data center.

The most valuable feature is its exceptional performance and storage efficiency. This efficiency translates to significant cost savings for us.

There is room for improvement in tier one support, especially with potential language barriers and communication challenges.

I have been using it for ten years.

We've experienced a few minor issues with stability, but we haven't encountered any significant outages.

Scalability has been a trouble-free experience. When we decided to increase our storage capacity by incorporating an additional drive shelf, the engineer collaborated with us, and the expansion process was executed smoothly, ensuring there was no downtime or interference with our production operations.

We rarely encounter performance issues, and we prefer to handle most of our troubleshooting internally. Our experience with tier-one support from NetApp has been somewhat lacking, and it varies depending on the support representative. We aim to resolve issues ourselves whenever possible, despite paying for support. On an overall scale of one to ten, I would rate the support as average.

Neutral

Over the years, we've tried various storage solutions like Nessus, Nimble, and IBM. However, about a year ago, we made the switch to NetApp, and we've been quite satisfied with their performance and have remained loyal to their products ever since.

The initial setup has been quite straightforward.

Cost reduction has been a significant benefit. We've been able to lower our expenses by minimizing the number of devices and drive shelves needed. Instead of having twenty drive shelves, we now only require two. This reduction in hardware has had a positive impact on power consumption, cooling, and maintenance costs, leading to overall cost savings.

We find the pricing to be favorable due to the educational sector we belong to.

Overall, I would rate it nine out of ten.

We use it to store all kinds of data, both structured and unstructured.

The way that it has helped our organization is that it requires less time to manage. It's almost like a set-it-and-forget-it type of solution. We don't have to do too much maintenance. Compared to other products, it doesn't need so much babysitting. It's easy to set up and it works. It does the things it is expected to do.

In addition, it provides unified storage no matter what kind of data you have. It has multi-protocol support. It does shares and it does block, so it's a one-stop solution that can fit all of your needs. You don't need multiple solutions for your different types of data.

The most valuable features are that it's 

• reliable
• simple 
• performs well.

It also helps to keep control of storage costs.

The only issue I can think of is metrics, but I think they have improved that in the newer versions already. There should be an easy place to see all your metrics.

I've been using NetApp Cloud Volumes ONTAP for more than 10 years.

It's very stable. We haven't had any issues since setting it up. It all depends upon the disk. The faster it is, the better the performance.

It is a very scalable solution. We are looking at how we can grow in the cloud and it can definitely scale in the cloud.

The support is good in general but the initial, frontline support could be improved. Because I have already been using the product for so long, when I call support I would rather talk to somebody who is a little bit more advanced or senior, rather than talking to the first-level support. Usually, it takes some time to reach out to their senior support. The advanced support is good. The frontline support can still be improved.

Positive

We've used Dell EMC in the past. We switched because NetApp definitely provides us with multi-protocol support and it is a one-stop solution.

It's fairly easy to set up. For a new SAN it takes a couple of hours to get the setup done. The additional configurations take another three or four hours. You can get the whole thing, a new system, set up within a day so that it is ready to go to testing.

Our implementation strategy is that we use CIFS shares and NFS shares in our environment. We also have block storage for SQL and Oracle. That has been our general plan all along. We separate these protocols by virtual servers. Once the necessary cabling is done, it's a matter of setting up the network interfaces for each, provisioning some storage, and testing things out. Overall, it's fairly straightforward.

I strongly recommend the solution. The biggest advantage is that it works as expected. There's less maintenance so you don't need too many people to support it and you save money in the long run.

We are predominantly using it as a backup target for our products. We are also doing some CIFS shares to remote sites that don't have their own file server infrastructures.

It gives us flexibility. In a disaster situation, or even in an office relocation, there can be a gap. NetApp CVO allows us to continue to provide service customers with access to their data, even if a physical site is going to be down for a long period of time. It's only really viable if you know a site is going to be down for a long period of time. We've had office relocations and there have been gaps between when the old office closed and the new office opened, during that period of moving stuff over and setting things up. There were a couple of weeks where we were serving the data out of the cloud, rather than out of the physical site. NetApp CVO may have improved our uptime by 1 or 2 percent, because we don't have that much downtime to start with.

It has all the advantages of the real NetApp product. You can provide storage in most of the formats you'd want. 

It helps us to keep control of storage costs because it's an OpEx-based model rather than a CapEx-based model. It depends on how you license it. You can have it up and down, almost on an hourly basis. Obviously, we don't do that, we've got it up long-term. But it does have that flexibility to bring up an instance of a client filer for just a short period of time.

It has saved us from having to buy and host another filer somewhere. That would be the only option to achieve the same goal. If we were to buy another filer to provision the capacity we've got in the cloud, the CapEx would probably be at least $200,000, whereas the running costs are not that much. It depends on how you deal with AWS, but we don't pay that kind of money. It probably saves us 75 percent of the cost of buying a filer for real.

One of the most valuable features is its similarity to the physical app, which makes it familiar. It's almost identical to a real NetApp, which means you can run all of the associated NetApp processes and services with it. Otherwise, we would definitely have to deploy some hardware on a site somewhere, which could be a challenge in terms of CapEx. Also, in our case, in Europe, in terms of physical real estate, we are trying to reduce the size of our data centers.

There is room for improvement with the capacity. There's a very hard limit to how many disks you can have and how much space you can have. That is something they should work to fix, because it's limiting. Right now, the limit is about 360 terabytes or 36 disks.

I've been using NetApp Cloud Volumes ONTAP for about two years.

The stability has been really good. I don't think we've ever had any major outages. AWS, obviously, doesn't guarantee 100 percent uptime, so I can see that it's not been up since I last restarted it. Rather, it's been up since some AWS event resulted in it migrating to another one of their pieces of hardware. But we've never had it actually crash.

The scalability is good to a point, but there is a hard limit on the capacity. We could, obviously, create another associated instance of it, but it wouldn't be a single name space, and we couldn't do some of the things you can do if you have a lot of multiple, real NetApps. So there are some hard limits to how big a solution you can create.

Day-to-day, it's probably only being used by about a dozen people in our organization, because it is mainly a backup target. There is a small collection of people whose shares live on it, but the majority of the business' files are on the real NetApps on their sites.

It's probably at a size where we're not likely to implement any more. You never know. It's very hard to tell what will go on with our company. But at the moment, it's probably not going to get any larger. We may actually shrink the capacity because we are temporarily storing some stuff for a part of the business that should only be on there for a few months at most, with this COVID.

As an organization, we went ahead wholehearted that anything and everything should be in the cloud — cloud first — and that got tempered a little bit because they started to see the costs. We also hit limitations with some of the software vendors because they're quite small companies and very niche. They don't want to support anything that's in the cloud, so there are limits to what you can put in the cloud.

Their technical support is very good. In the early stages, we would get almost instant online support, because we would go into the Cloud Manager and there would be a chat and we could have a chat session with the engineers who were implementing it on the NetApp side.

As things have progressed, we now need to follow a more formal support model, but we usually get a pretty good response, for general, routine questions, within five or six hours. If it were a major incident, you would get much faster support. We've never had a major incident with it.

It replaced some physical NetApps that were going to be refreshed. One of the reasons we switched was to limit capital expenditure. Another reason was that it was very much a "Let's go and put as much as we possibly can into the cloud" approach. It fell in with that initiative quite well.

The initial setup was pretty straightforward. The challenges we had were only around the security we put on top of AWS. For me, as an engineer, to be able to do things requires another team to do stuff on the network side or to do stuff on my rights within AWS so that I could deploy it and manage it afterwards. But it is relatively straightforward if you're not fighting other complications.

It took us a couple of days to get it up and working the first time. My colleague did one in the US and it took him about half a day. We did one for another part of the business and that took about three or four hours to get up and running.

Initially, we were just doing an evaluation to see what it was like and if we could actually use it. It went from a trial implementation to going live within a month or two, once we realized it was going to do what we wanted to do.

We had four people involved in the implementation. I was involved, as a storage engineer, and we also had one of our client specialists, a network person, and an info-sec person to validate that the network stuff was within their rules. In terms of maintenance, it's just  me, but it doesn't really require a lot of attention because it's cloud-based and it's a NetApp. Generally, once you set them up properly, unless you're changing something, they look after themselves.

It was done by just us. Because it was one of the very early implementations of Cloud Volumes ONTAP, we were working with NetApp and their staff were playing the role that a third-party integrator might have played.

We're probably burning about $10,000 a month on it but it's saving us the CapEx and the power and cooling of a real filer. We're likely seeing at least a 50 percent saving.

Choose your disk type properly. Go with the slowest, cheapest disk you can. If you need bigger, faster ones then go for them. 

They've got a variety of license schemes. The one we've gone for is where we pay NetApp once a year. They call it the Bring Your Own license scheme. There is a by-the-hour or by-the-month basis from AWS and you can get it that way as well and be billed through AWS. But you may not get the same level of discounts that you would if you were dealing with NetApp directly. If you are committed to having a client filer for an extended period, then go with the NetApp licensing model rather than the AWS-provisioned one.

Ultimately, the more data you save, the more it costs you, because you're paying AWS for the capacity. NetApp is licensed per filer, but there are additional running costs that are paid to AWS. You pay AWS' hosting fee for an EC2 instance, and each one of the disks within the NetApp is EBS storage and you pay AWS for those.

There is potential to save money by moving things off to object storage. The only cost savings we see on it is against having to buy physical hardware.

We looked at third-party hosting with either our own, dedicated hardware or shared NetApp hardware. I wasn't that involved in that evaluation process, but I figure that the costs for the work-around were too high or the solution was too complex for us to go with.

CVO enables us to manage our native cloud storage better than if we used management options provided by the native cloud service. With the native solutions, you don't get any of the advantages of the NetApp in terms of being able to deduplicate and having clear management of the snapshot-ing. Also, at the time, there wasn't an easy way to back up to a cloud NetApp. There was nothing. Now they have a slightly different solution where they'll mount it for you but, at that time, you created your own cloud instance and your own cloud file and you managed that. Now, you can access a solution that is managed by AWS or by NetApp.

It is almost identical to having a real NetApp, and it's just that it's remote and it's in the cloud. Almost anything you can do with NetApp locally you can do with a cloud filer.

Go with the cheapest disks to start with, and if you need the performance you can easily transition to using faster disks.

There are limitations, but in general it's robust and easily managed.