NetApp Cloud Volumes ONTAP Review

High availability enables us to run two instances so there is no downtime when we do maintenance


What is our primary use case?

The primary use case is for SAP production environments. We are running the shared file systems for our SAP systems on it.

How has it helped my organization?

It's helped us to dive into the cloud very fast. We didn't have to change any automations which we already had. We didn't have to change any processes we already had. We were able to adopt it very fast. It was a huge benefit for us to use the same concepts in the cloud as we do on-premise. We're running our environment very efficiently, and it was very helpful that our staff, our operators, didn't have to learn new systems. They have the same processes, all the same knowledge they had before. It was very easy and fast.

We did a comparison, of course, and it was cheaper to have Cloud Volumes ONTAP running with the deduplication and compression, compared to storing everything, for example, on HA disks and have a server running all the time as well. And that was not even for the biggest environment.

The data tiering saves us money because it offloads all the code data to the Blob Storage. However, we use the HA version and data tiering just came to HA with version 9.6 and we are not on 9.6 in our production environment. It's still on RC, the pre-release, and not on GA release. In our testing we have seen that it saves a lot of money, but our production systems are not there yet.

What is most valuable?

The high availability of the service is a valuable feature. We use the HA version to run two instances. That way there is no downtime for our services when we do any maintenance on the system itself.

For normal upgrades or updates of the system - updates for security fixes, for example - it helps that the systems and that the service itself stay online. For one of our customers, we have 20 systems attached and if we had to ride that customer all the time and say, "Oh, sorry, we have to take your 20 systems down just because we have to do maintenance on your shared file systems," he would not be amused. So that's really a huge benefit.

And there are the usual NetApp benefits we have had over the last ten years or so, like snapshotting, cloning, and deduplication and compression which make it space-efficient on the cloud as well. We've been taking advantage of the data protection provided by the snapshot feature for many years in our on-prem storage systems. We find it very good. And we offload those snapshots as well to other instances, or to other storage systems.

The provisioning capability was challenging the first time we used it. You have to find the right way to deploy but, after the first and second try, it was very easy to automate for us. We are highly automated in our environment so we use the REST API for deployment. We completely deploy the Cloud Volumes ONTAP instance itself automatically, when we have a new customer. Similarly, deployment on the Cloud Volumes ONTAP for the Volumes and access to the Cloud Volumes ONTAP instance are automated as well.

But for that, we still use our on-premise automations with WFA (Workflow Automation). NetApp has a tool which simplifies the automation of NetApp storage systems. We use the same automation for the Cloud Volumes ONTAP instances as we do for our on-premise storage systems. There's no difference, at the end of the day, from the operating system standpoint.

In addition, NetApp's Cloud Manager automation capabilities are very good because, again, it's REST-API-driven, so we can completely automate everything. It has a good overview if you want to just have a look into your environment as well. It's pretty good.

Another feature which gets a lot of attention in our environment is the File Services Solutions in the cloud, because it's a completely, fully-managed service. We don't have to take care of any updates, upgrades, or configurations. We're just using it, deploying volumes and using them. We see that, in some way, as being the future of storage services, for us at least: completely managed.

What needs improvement?

Scale-up and scale-out could be improved. It would be interesting to have multiple HA pairs on one cluster, for example, or to increase the single instances more, from a performance perspective. It would be good to get more performance out of a single HA pair. My guess is that those will be the next challenges they have to face.

One difficulty is that it has no SAP HANA certification. The asset performance restrictions create challenges with the infrastructure underneath: The disks and stuff like that often have lower latencies than SAP HANA itself has to have. That was something of a challenge for us: where to use HA disks and where to use Cloud Volumes ONTAP in that environment, instead of just using Cloud Volumes ONTAP.

For how long have I used the solution?

We've been using Cloud Volumes for over a year now.

What do I think about the stability of the solution?

The stability is very good. We haven't had any outages.

What do I think about the scalability of the solution?

Right now, the scalability is sufficient in what it provides for us, but we can see that our customer environments are growing. We can see that it will reach its performance end in around a year or so. They will have to evolve or create some performance improvements or build some scale-up/scale-out capabilities into it.

In terms of increasing our usage, the tiering will be definitely used in production as soon as its GA for Azur. They're already playing with the Ultra SSDs, for performance improvements on the storage system itself. As soon as they become generally available by Microsoft, that will probably a feature we'll go to.

As for end-users, for us they are our customers. But the customers have several hundred or 1,000 users on the system. I don't really know how many end-users are ultimately using it, but we have about ten customers.

How are customer service and technical support?

Technical support has been very good. The technical people who are responsible for us at NetApp are very good. If we contact them we get direct feedback. We often have direct contact, in our case at least, to the engineers as well. We have direct contacts with NetApp in Tel Aviv.

It's worth mentioning that when we started with Cloud Volumes ONTAP in the past, we did an architecture workshop with them in Tel Aviv, to tell them what our deployments look like in our on-premise environment, and to figure out what possibilities Cloud Volumes ONTAP could provide to us as a service provider. What else could we do on it, other than just running several services? For example: disaster recovery or doing our backups. We did that at a very early stage in the process.

If you previously used a different solution, which one did you use and why did you switch?

We only used native Azure services. We went with Cloud Volumes ONTAP because it was a natural extension of our NetApp products. We have a huge on-premise storage environment from NetApp and we have been familiar with all the benefits from these storage systems for several years. We wanted to have all the benefits in the cloud, the same as we have on-premise. That's why we evaluated it, and we're in a very early stage with it.

How was the initial setup?

To say the initial setup was complex is too strong. We had to look into it and find the right way to do it. It wasn't that complex, it was just a matter of understanding what was supported and what was not from the SAP side. But as soon as we figured that out, it was very straightforward to figure out how to build our environment.

We had an implementation strategy: Determining what SAP systems and what services we would like to deploy in the cloud. Our strategy was that if Cloud Volumes ONTAP made sense in any use case, we would want to use it because it's, again, highly automated and we could use it with our scripting already. Then we had to look at what is supported by SAP itself. We mixed that together in the end and that gave us our concept.

Our initial deployment took one to two weeks, maximum. It required two people, in total, but it was a mixture of SAP and storage colleagues. In terms of maintenance, it doesn't take any additional people than we already have for our on-premise environment. There was no additional headcount for the cloud environment. It's the same operating team and the same people managing Cloud Volumes ONTAP as well as our on-premise storage systems. It requires almost no maintenance. It just runs and we don't have to take care of updating it every two months or so for security reasons.

What about the implementation team?

We didn't use a third-party.

What was our ROI?

We have seen return on investment but I don't have the numbers. 

What's my experience with pricing, setup cost, and licensing?

The standard pricing is online. Pricing depends. If you're using the PayGo model, then it's just the normal costs on the Microsoft page. If you're using Bring Your Own License, which is what we're doing, then you get with your sales contact at NetApp and start figuring out what price is the best, in the end, for your company. We have an Enterprise Agreement or something similar to that. So we get a different price for it.

In terms of additional costs beyond the standard licensing fees, you have to run instances in Azure, virtual machines and disks. You still have to pay for the Azure disks, and Blob Storage if you're using tiering. What's also important is to know is the network bandwidth. That was the most complicated part in our project, to figure out how much data would be streamed out of our data center into the cloud and how much data would have to be sent back into our data center. It's more challenging than if you have a customer who is running only in Azure. It can be expensive if you don't have an eye on it.

Which other solutions did I evaluate?

We have a single-vendor strategy.

What other advice do I have?

Don't be afraid of granting permissions because that's one of the most complex parts, but that's Azure. As soon as you've done that, it's easy and straightforward. When you do it the first time you'll think, "Oh, why is it so complicated?" That's native Azure.

The biggest lesson I've learned from using Cloud Volumes ONTAP is that from an optimization standpoint, our on-premise instance was a lot more complex than it had to be. That's was a big lesson because Cloud Volumes ONTAP is a very easy, light, wide service. You just use it and it doesn't require that much configuring. You can just use the standards which come from NetApp and that was something we didn't do with our on-premise environment.

In terms of disaster recovery, we have not used Cloud Volumes ONTAP in production yet. We've tested it to see if we could adopt Cloud Volumes ONTAP for that scenario, to migrate all our offloads or all our storage footprint we have on-premise to Cloud Volumes ONTAP. We're still evaluating it. We've done a lot of cost-comparison, which looks pretty good. But we are still facing a little technical problem because we're a CSP (cloud service provider). We're on the way to having Microsoft fix that. It's a Microsoft issue, not a NetApp Cloud Volumes ONTAP issue.

I would rate the solution at eight out of ten. There are improvements they need to make for scale-up and scale-out.

Disclosure: IT Central Station contacted the reviewer to collect the review and to validate authenticity. The reviewer was referred by the vendor, but the review is not subject to editing or approval by the vendor.
2 visitors found this review helpful
Add a Comment
Guest
Sign Up with Email