NetApp HCI Review

Separates control of storage from compute and enables the ability to individually choose the minimum and maximum burst on IOPS


What is our primary use case?

Our primary use for the product is doing demo presentations of it for customers. We show them the function of the appliance, the features, how easy it is to operate, and things like that. That is what our main purpose for it is as resellers.

How has it helped my organization?

The product has improved our organization by making it very easy to explain the product to a customer to show what kind of environment that they can use this appliance in. 

What is most valuable?

One of the most valuable features of the solution is that they have taken something like 400 steps and compiled it into like 30 steps to install it. That level of simplification is one of the best things I've ever seen in any appliance in the market.

Normally when you build an appliance, you have to go through an extended installation process. But once you collect all the IP's that you need to for this appliance, the NetApp engine does all the rest of the work for you. So instead of leaving it on the operator to do all the work, the appliance does it. That is one of the features that I have seen in any appliance that really simplifies a lot of the work you would otherwise have to do for setup.

What needs improvement?

An improvement I would like to see is if they could make it even more simple to set up. They were able to figure out a way to get 400 steps down to about 30 steps. Trying to reduce that down more would be an even greater benefit. 

What do I think about the stability of the solution?

Once the product is up and running with the build configured correctly, the product will run itself. It is highly stable.

What do I think about the scalability of the solution?

It is scalable. You could add more nodes. You have to start out with the minimum number of storage nodes. But then you could scale up as much as you want on just storage or just compute.

How was the initial setup?

The set up is pretty straightforward. If it does have any sort of learning curve the only thing is you have to be sure to get all the pre-checks done. I would say that has too many IPS (Intrusion Prevention System).

So that is a pre-check you have to do now and hope that they address in the future to somehow downsize that into a limited number of IPS. Right now it is like 16 to 24 IPS, which is a lot for one appliance.

What was our ROI?

The solution has reduced TCO (Total Cost of Ownership) for some customers — or really most. Because there is a hyper-converged infrastructure, you do not need to have that three-tier architecture with separate storage, separate compute, and separate network. It definitely has reduced costs. I would say at least by 50 percent in most cases.

Which other solutions did I evaluate?

The products that clients are usually evaluating are other hyper-convergent tools because NetApp HCI is a hyper-converged infrastructure. There are other parts of the market like Nutanics, VxRail, and Cisco (HyperFlex). When we help clients do a comparison of other products, NetApp wins because of the leading features. For example, they are the only product on the market that has independent compute and storage capability. No other the product has that.

What other advice do I have?

Because our organization functions as resellers, we use the product features in demonstrations. We do not actually apply them to our own business. For example,
the solution's ability to scale on-demand and affect provisioning in an organization is something we demonstrate to those who are interested in the product. We do not use it personally. We explain that feature in our presentation and the customers want to see it. When they do they are wowed and want the capability. 

This solution's ability to scale compute and storage independently affect capacity, performance, and operational planning in our customers' organizations. The best thing about the product is this is the only appliance in a market that separates control of storage from compute. There is nobody in another OEM (Original Equipment Manufacturer) on the market that can do that currently. It is absolutely unique because you can individually choose the minimum and maximum burst on IOPS (Input/Output Operations Per Second). Nobody lets you do that independently like this appliance does.

The solution also has the capacity to reduce your hypervisor footprint. We do not actually have that many customers that use hypervisor because most everybody is in the VM world right now. It is a plugin in vCenter and it a very usable, very friendly tool as a plugin.

Being that NetApp HCI is a hyper-converged infrastructure a company with a data center will not have to have separate storage, special computers, and a special network for different purposes. This product does all of that in one central appliance. Which is why HCI is very popular.

The solution increases application performance. Because it works well with VM products and lets you independently choose how much of the resources each VM could use as far as IOPS. There is no noisy neighbor — no one appliance or application gets to hog the VM bandwidth and resources. So this one you can separate those roles out and silo it to make it a lot easier.

The solution affects the storage performance for organizations as well. We only work with it in a demo presentation environment and we do not have production data so the demos are just a sample of workloads. But no customers that we have come across have any issues and the pain points with storage performance.

Results show more efficient use of compute resources. Again, being the only appliance in a market that lets you independently choose compute and storage, you scale as you want. If you want more compute you can add more compute nodes. You want more storage, you add more. It simplifies a lot of things.

On a scale of one to ten where ten is the best, I would rate the product as a nine. The reason I give it a nine instead of a ten is because of the need for pre-check work. While it is simpler than most products, there is still a lot of configuration that needs to be done. If they make that simple it improves the product. When you have to rebuild the product, you have to go through that whole process again. If they could somehow make that simpler by keeping that configuration file in a way that you do not need to do that, we could just use that configuration file again. It should automatically rebuild itself instead of you going through the whole process another time.

Advice that I would give to a colleague who is researching this is to use the product. Do a proof of concept and that will open up the customer's eyes to imagine the possibilities and see how this product of NetApp works. If I was in that position, I would push this product more to show a customer proof of concept, show it and the environment, and then they would tell the reseller to leave it there because they will want to buy it.

Which deployment model are you using for this solution?

On-premises

Disclosure: My company has a business relationship with this vendor other than being a customer: Partner.

Add a Comment
Guest