NetApp AFF - All Flash Reviews

It is cost-effective flash for us. It's a platform that we've used for quite a few years. We've been NetApp customers for probably about eight years right now. You don't have to go in and re-learn any kind of new interface; it's using basically the same interface. Provisioning is ultra-fast and it just works.

We had a few databases that were using SQL databases that we were having some performance issues with. We moved them over to FAS and, no more performance issues. Basically, you throw a ton of hardware at a problem and that fixes it.

A lot of the applications that we use are canned applications. We don't actually have the ability to go in and modify them. We’re kind of handed a bad deal in some aspects. We go in, we put that stuff on flash to see whether we can make this thing perform the way it's supposed to. We really don't have the option of going in and changing the code.

I recently attended a conference and one of the sessions was about performance data with ONTAP 9. They've addressed some of the issues that I'd like to see in that, such as being able to see where your latency is, and how much performance you have left in the array before you need to start looking at what we need to start moving workloads around.

It would be a little bit nice if the monitoring was a little bit better and smoother, but we've not had any issues from that perspective. In the future, I don't want to have any of those issues.

We've had no stability issues so far with it; in fact, I've not lost a disk out of it.

Scalability is not something that we're going to be concerned with right now, as far as adding; we can always add a tray. It's non-disruptive. That's great.

For flash, we did use professional services to come in and help us get it set up but other than that, we've not had to make any phone calls about it. It's pretty straightforward.

At the time, we had been NetApp customers for quite some time. We had been using a FAS3220 and we were starting to see performance issues. Our sales engineer said, “Why don't you guys try take a look at this?” We did some research on it. We actually POC'd it with a few others, that will probably remain nameless at this point in time, and obviously, NetApp outperformed the others; oh, we loved that.

Well we were a 7-mode shop and we were switching to CDOT. There's a little bit of a transition there. I won't say it's overly complicated; it's just some new things to get used to.

The setup complexity is why I have not given it a perfect rating; not that it's a big deal. We had professional services come and help us do that but going from 7-mode to CDOT was quite a jump.

I have a feeling that’s pretty common. We were going through the conversion on all of our arrays. We currently have three. It's getting easier as we go through the process more and I understand it. It would be a lot better if the transition was a lot more smoother.

A lot of companies will tell you that they're the best at what they do. As a company, I think it's very important that you look at POCs to see if you can get them. Everybody can tell you they have the best product, but until you can actually prove it on your workload, you really don't know 100% for sure.

When selecting a vendor to work with, as a company, we have had a tendency to just go and buy the "best of breed," which sometimes included arrays from multiple vendors. As a company, we have five different brands of arrays. You can't become an expert in something if you have five different arrays to work with. What we're trying to do as a company is to align to say, best of breed being, this is fantastic as a NAS appliance so we're going to look at that and say, maybe we should look towards getting that. I think we're taking our shotgun approach and we're kind of moving it down to where you can be more specialized in what you do. As I’ve mentioned, NetApp is fantastic; it does block, it does NAS. It's a one-stop shop.

For us, the most valuable features were the SnapMirroring, deduplication, and inline compression. Now with 9.0, the compaction system, that's actually the big thing that sold us on it besides just the price in general. It was a very well-priced system for what we got. The data dedupe and inline, we're getting substantial rates. I think it's about 60-65% in general. That's a massive savings over what you would get if you didn't have any of that stuff.

We have a job system that runs all the time; people can run what they call campaigns. It drastically increased performance. It decreased times by three times the amount. The amount of the CIFS shares increased from about 128 Mbps – it was only a 1-gig line anyway, to a 10 gig – to about 800 Mbps. The engine actually can't pull enough and it has caused a little issue here and there, because it's basically causing a race condition. We've had to program around race conditions because we haven't had a system that was this fast.

It saved us a lot of time as well, substantial savings.

If they could do the tabbing for the nodes, that would be spectacular. On 9, they offered more insight, so I can't really say that. We haven't upgraded both nodes. We have HA pairs, and one of them is still running 8.3.2. We upgraded our DR solution to nine first just to see if it causes any issues. So far, we haven't seen anything. They have a lot more insight into that; I wish they would have it on 8.3.2 but, what are you going to do?

So far, stability’s been excellent, and the update process was actually incredibly painless. We've upgraded twice now and I am surprised that it didn't cause any issues at all. Usually, you have to have some kind of user intervention. For this product, you just throw the image on there, click update and it's done. You come back about an hour later and you're happy.

The GUI is really good, but if you don't find the option in the GUI, then the CLI is amazing. You can hit Tab and just tab out. The only thing is, they haven't done that on the nodes themselves but I was told they're working on that.

We don't have that big of one yet. We originally quoted out a system of eight nodes, and it was going to be something like 12 GBps. That seemed like substantial amounts, considering what everyone else quoted. However, it actually was going to come in at about the same price for the AFF compared to everyone else's quotes for disks. The reason they went with it is because of the trust with the vendor they were currently using and they just didn't want to leave.

We were using NetApp before, but we evaluated EMC, IBM, HP, Pure Storage, XtremIO and Nimble.

It came down to XtremIO and NetApp. NetApp offered much, much more storage. And the cost difference to buy XtremeIO was huge compared to NetApp. NetApp just totally blew it out of the water on price. We got something like five times the storage for the price. It was really worth testing on that.

Try out what you actually want to do, because that's actually the problem we had; some of our people swore up and down that NetApp wouldn't be able to do compression at the new rates that they got, or that we got. They said that Oracle doesn't compress and so on. We ended up getting them to stick some of their machines on our NetApp, and we showed them that you actually do get it.

We actually bought ours and then we tried to show those other people before they got to the bidding table for theirs. They didn't really want to listen to the facts. They went with IBM. I wouldn't say they were not unhappy or anything. They realized that they could've gotten a lot more if they just went with our ideas instead of their idea. Actually, I was told it was more of a management thing; they actually didn't even want IBM, they wanted Oracle. It all comes down to what the boss says.

• Speed
• Performance
• Low latency

We do not use it now for our customers, but just to manage our criticla internal CRM proposals. It helps us manage our entire CRM. We had problems with latency in the regular NetApp storage devices (many concurent access attemps for small data) and AFF has improved it.

The price needs to come down. Also, it has a learning curve and I needed to learn a lot to do the installation.

We had no problems with deployment.

From what I know, no problem at all with stability.

We have not scaled it yet, but we are thinking about it.

Customer service is perfect every time.

Technical support helps us every time.

We used IBM before, after the contract finished we tested NetApp, and from what I know it filled the hole. It’s excellent. It has flash pool disks on a cluster, and we switched as we got it for a good price. We have a dual strategy with Inter Telecom and normally we get good prices for NetApp products.

I set it up the first time myself. I needed to learn it and read the user guide for it, but it’s easy for technical people.

We tested many products, but no other flash ones.

I would advise you to make sure that you need flash as it is very specific and regular FAS may work for you. However, if you need flash, this is a good product to get.

The most valuable feature for us completely depends on the workload. We've run it on various environments, including VMware. We were able to migrate to VMware with clustered Data ONTAP. That was important for us.

Stability was something that was really wanted to improve on, and we've been ale to do that with AFF. Previously, we had to reprogram every three months or so, so not having to do that with AFF is huge.

Secondly, we do a lot less daily maintenance than we used to. It's a fairly trouble-free system as we simply configure it and pretty much leave it alone.

I had a few issues, but they were easy to fix. I would like for it to be more scalable as the number of shelves is limited.

We installed it in July.

Scalability is difficult. The number of shelves is limited to 2 or 4, and the number of terrabytes we potentially have doesn't match to this limit.

We don't have a lot of contact with technical support because our engineers handle any issues. But the product itself has been excellent.

The initial setup wasn't easy, and in fact took about 6 years. It was a slow start, but of course it was new back then, and it takes time.

We were one of the first and even the engineer had some trouble configuring it for us.

We can't decide for ourselves so we ask the market, this is what we want to do, what should we buy? Then markers come with the products. There isn't a lot of choice.

We've had a good experience with it so far. Right now, we're using it for our customers, but if we were to also use it for ourselves, it would be too small. I'm sure they'll improve in the future, but we'll just have to wait for the solution to support ten to fifteen thousand users.

We use it for typical data center workloads: Exchange, file shares, and SQL.

We have a big problem in our organization where I can't get the application engineers to give me performance requirements. Now, with the SSDs, I don't need to worry about that anymore. All of our applications are high. Our test applications perform at a higher level now.

It has improved performance of our enterprise applications, data analytics, and VMs because we have a higher IO from the disk now. We run a lot of write-intensive VMs. For sure the solution helps out.

Our total cost of ownership has decreased because of the nature of the SSDs, their mean time to failure is much higher. They don't fail as often and that's going to reduce it. And because we upgraded to the All Flash and the bigger SSD, we reduced our footprint. I increased my capacity 500 percent and reduced my footprint in the data center by 95 percent.

The most valuable features are

• the IO performance that we get
• the cluster part 
• the increased workload and performance with the SSDs.

And the CLI portion of ONTAP, in general, is much easier to use.

It's a little behind on security. It's starting to get into multi-factor authentication, they just started to introduce it but not for all products. In my area, we are really big on security, using smart-card authentication. Multi-factor authentication is a big thing for us, being on the federal government side of things. We need all the products to have the ability to do smart-card authentication. That's the biggest one. That's the drawback of this solution. But otherwise, it's getting there. It's starting to catch up.

It has been very stable so far. It's about a year old, we haven't been using it for long, but so far it has stood up very well.

We haven't needed to scale it yet. We probably won't. But obviously, because we are in a multi-node cluster environment, with the switches we can scale out very easily if we need to.

I mostly interact with my sales engineer who is very sharp. The few times that I've had to interact with technical support, it has been very good.

The gear we were on was about ten years old. We always buy behind the technology curve. I noticed that spinning disk was going away and that the industry moving towards SSDs, so I wanted us to try to get ahead of the curve a little bit, to give us some more horsepower to do some more initiatives that we want to get done in the future.

It was very straightforward. There are setup tools so if you're not very familiar with NetApp, they walk you through the process step by step: How to configure all the interfaces and the SVMs, etc. I'm more experienced with the command lines, so I deployed it that way. But it's very receptive to PowerShell scripting, so it's easy to use.

We used an integrator, reseller, and consultant for the deployment. Resellers are resellers. I don't have a good or bad opinion of them. As for the integrators we had, I'd rather do it myself quite honestly. But it was okay.

Because we're federal government, we really can't choose. We've had NetApp for years. I did evaluate a lot of other products. Honestly, at the end of the day, storage is storage and disks are disks; it's all the bells and whistles on the front. Other solutions could probably have accomplished the same task. Ultimately, it comes down to dollars and cents, but I'm not really involved in that side of it. I'm sure they chose NetApp because of the cost.

Know your workload, know your customer. Know what your requirements are, know what your future requirements are. Determine what's important to you. Think about the administrators, if you're not the administrator; I'm not, I just engineer it. Think about them and how they will use it. Think about the future, where you think your business will grow.

When it comes to setting up and provisioning applications using the product, it depends on what you're doing. But I I can have an Exchange server up and running in about 30 minutes.

At the moment the solution is not having any effect on IT's ability to support new business initiatives. I got it to support things like ADI and solutions like that. So hopefully, going forward, it will play a role in that. We have not connected the solution to public clouds. We do plan to in the future.

I rate the solution an eight out of ten because there's room to improve. There's always room to grow. The security side of it: They have a large government customer base but it seems like they really don't pay attention to that side of things. There are a lot of security things, a lot of customers can't send their stuff offsite, and I'm one of them. So coming up with better ways to satisfy that part would be great.

• Performance
• Block storage

It improves organizational performance.

• I want to see more send features.
• It takes awhile to learn the system.

It is stable.

It does not matter much in our environment. We have not thought of scaling out.

We have used the technical support. They are good.

We had EMC, then we introduced NetApp. We switched due to cost.

I was involved in the initial setup. There was a little bit of an issue, but it turned out okay. Basically, we had to call NetApp for assistance during the setup due to an odd issue.

We use AFF as part of a cluster with other NetApp class systems.

I would definitely recommend AFF.

There's always a little room for improvement.

We have had no issues, but we never went to production.

The technical support is excellent. Anytime we've had any kind of questions, our rep can help us or we'll call into NetApp auto-support. We have not had any problems. Tech support is knowledgeable and their response times are good.

We compared this tool against EMC’s XtremIO head-to-head, and the NetApp blew it out of the water. There was no competition. We were already a NetApp shop, so they were our preferred tool anyway. It has more features and links to my OS, innovative CIFS, and deduplication. We had the knowledge of the system already. It wasn't reinventing the workforce.

The installation was pretty easy. It was my third setup, so it was nothing really new. There's only one minor switch that turns it into an AFF.

We use the system to do stuff that isn't quite out yet. We love to do some oddball things. We're one of the first to use NetApp shift to compete and migrate away from VMware. We didn’t run into any issues with it, and it beat the competition.

When looking for a vendor, it's usually value first, which is not the right way to do it. That's what it comes down to. The value and then next is feature set.

Dedupe (cost saving): We were able to achieve a lot more capacity than expected.

• More desktops on storage
• Ease of management

• Software packaging and ordering.
• We wanted to integrate with replication and Commvault options, and that was difficult.

We have been using the solution for about six months.

We have not encountered any stability issues.

We have not yet encountered any scalability issues.

I would rate the technical support at about 8/10.

Other solutions were not all-flash compatible.

Initial setup was easy.

Compare and look for your use case.

We evaluated Pure Storage, SolidFire, EMC Unity.

The migration plan should be clear upfront.