NetApp EF-Series All Flash Arrays Reviews

We're able to quickly roll out close to 2,000 VDIs with very little setup. I've had familiarity with a different model of the NetApp device. It was mostly a matter of rolling it over to the new one. It was very easy to work with. The familiarity with it and the speed of it; I can't complain.

The most valuable feature is its speed. We had a bake off with another product in our data center, and it blew it away, without any question. It was about 100% faster than the XtremIO. It was just pure speed.

We use it for our VDI environment, and have not had any complaints with it.

At this point, we haven't really gotten that far in our stages of it. The scalability is what we need. The administration is really easy. The best way I can say this is, keep on keeping on. It's going to grow organically.

If anything else, I would like to see higher-capacity drives, as they come out; I have heard that 15 TB are out on a different NetApp series. Getting those on the EFs would be nice. That's maybe a pipe dream. The EF series doesn't have them quite yet. I don't think they're really designed yet for the EF series, from what I’ve heard from a couple of solutions engineers. If that happens, great; if it doesn't, great.

I have not had any stability issues with it. We actually have two shelves for over a year now. It does not give us any problems. We're not even really pushing it either. That's kind of the other good side of it. It's got a lot of capacity, a lot of IO availability, and bandwidth. It just works.

Everything we've wanted to do, scale-wise, it's done. We started off with one shelf, got about 2,000 VDIs on it; put another shelf in, not a problem. It was about as easy as you can make it.

Unfortunately, we have not used technical support. It hasn't had an issue for us to do that, yet. In the 15 months that I've been there, I haven't had to call support for that product.

XtremIO is a nice product, but it has so much more overhead. Deduplication involves a little bit more overhead than what the EF does. I kind of look at EF as a dragster as opposed to a road course racer. The EF just goes straight line. It does what it's supposed to do; not a lot of frills, but it does its job really well. XtremIO is a bit more tuned for a general-purpose workload; not so much speed. It's still flash, but the overhead on the XtremIO was more than what we need it for. The price point was better on the EF. You get more speed, better capacity, and better price. The EF was a better value.

The disadvantages of the XtremIO were the speed and the capacity. I think it was a very similar scale out for the XtremIO. It was very similar in size. It was a little bit smaller because there was more overhead; different metadata for the XtremIO. It required a little bit more controller size, and so on.

Also, it was a 4U brick versus a 2U brick. We basically get twice the capacity of two shelves for NetApp versus one brick from XtremIO. The scalability was one of the big features as well.

When selecting a vendor to work with, the one that provides the best value is the most important criteria; does it hit the most amount of needs that we have; once we determine what our requirements are, and how we go forward. What's the best for us? We may spend a little bit different money on something else, but that has to be of a specific need that we need to fit. Basically, what's the best value?

As I’ve mentioned, from familiarity with a different version of the product, initial setup was very easy for us. My cohorts and partners in crime are very familiar with it. They have no problem looking at it, either, and understanding what's going on with it. It's been a very good product for us.

If you can get a bake off, do that. Try to get the same type of test across the board. Put it through its paces. Definitely get your solutions engineers involved; almost pit them against each other. Ask a lot of questions, and really find out what the requirements are. Get them in there and try them out. You lose nothing, except a little bit of time. If you can spend better money the first time out, you look better as an engineer and a person that can influence purchasing.

I have not given it a perfect rating because, as I’ve mentioned, it doesn’t have some features. At the same time, I have rated it high because it does its job really, really well; what we require of it.

More things get done faster. Time is money. If we have systems that are down for more than 10 minutes, that's $10,000 out the door, basically. They want true speed and being able to get up there.

The speed is the most valuable feature. It's a really good system. It's really fast; the speed's the best thing. It's one thing we love. Everybody wants to have everything faster and faster and faster. Knowing the speed's there, that's the best thing, along with the size and storage. I'm interested in the new stuff that's coming out, with the 32 terabytes. It will be interesting.

Once I play with it a bit more, I'll find out more about areas with room for improvement.

Cheaper pricing is always good. NetApp has been doing everything the right way. They've been figuring out things really well, going in the right direction.

I’m looking forward to the new shelves, the new disks and their sizes; how small you can get for lots of storage now. The size and speed are just amazing; always a good thing. Smaller systems save on electricity, save on space. We shrunk our server room down. We actually rebuilt it. We didn't have enough space for people to sit. We needed more office space. With everything so small, we cut our server room down and fit more people in. Our head count can go up and everything. People had jobs.

I would like to shrink it more, if we can. The smaller, the better.

I think stability is really good. We just started using them now. We're just getting into it. We're getting more and more systems as we go along. So far, everything has been stable; we have not had any problems. It's all new.

We never used flash arrays or anything like that before. This is all SSD and this stuff is all new to us.

We'll find out about scalability. I think it will be perfect. We have small sites. We just purchased a big site, a half-a-million dollar site, for Plymouth, Michigan. We'll find out how that works out, scaling up from that point. We used to do the shelves, the SATA shelves, the SAS shelves and so on. Flash is completely new and any SSD drives is all completely different to us. It will be interesting.

I haven't dealt with technical support yet.

We go off a price list. We kind of figure out what best fits the needs of each site. We have hundreds of sites in US and Canada.

We just wanted the speed. Everybody wants better IOPS and that was basically the reason why we chose it, to have better speed. Our sites need to be up; they need to be running.

We previously used regular SATA and SAS shelves, strictly disk shelves and so on; no flash, no arrays, nothing like that.

Before choosing this product, we did not really evaluate other options. They’ve looked at IBM and EMC. They've looked at that for some servers but most of the time, everything we stick to is with NetApp. We don't go all over; we stick strictly with NetApp.

So far, it's been fitting our needs. I've not had any issues with it, but it's all brand new to me. Right now, the install is great; everything is running fine; we haven't had any problems.

Go with the NetApp EF-series All Flash Arrays, if you want good quality, reliability, speed and size. I think that's the way to go nowadays; flash is amazing. I'd give it the thumbs up to do it. Spend the money. It might cost a little more but the quality is the best, for me. Sometimes, you have to spend money for good quality.

When selecting a vendor to work with, the most important factor for me is the relationship. We've had a great relationship with our sales managers and sales reps and we saw them at a recent conference. That's a key thing. You get the support you need. If something happens, they're on top of it, fixing it right away. Good service is the most important factor.

Cost: it's a lot cheaper. It's a lot cheaper than what the other vendors have for the same type of environment. It saves us money.

The ease of use is the most valuable feature; something that we can use for a flash and all-flash type of environment that we can really put a heavy workload on. It gives us an environment to where we can really push a lot harder. We have multiple vendors in our establishment. We have NetApp, EMC, IBM, HP; we have pretty much all of them.

We do a lot of in-house, application-dependent type things, where we find the different niches to the different things. Certain things they do better. We've found that it actually does very well on some of our higher-end applications.

I would like to see a way for NetApp personnel that are used to the FAS and the CDOT systems, to be able to easier translate that experience and knowledge into using the EF series. There are some differences in there and it will throw off, when you're trying to train somebody, as far as bringing in somebody new into the group. When they're supposed to be responsible, it's another technology for them to try to learn. Something that would help port that process; make them similar in how the manageability of it is functioned.

Obviously, everything can be improved on so I won't ever give anything a perfect rating. But as far as the manageability, being able to port between the two and have to do less training in-house from a customer point of view, that would be the part to improve.

We have not had an outage with it yet. Other vendors, we have. We've had an actual high-end frame completely go down on us.

We haven't done too much with scalability yet in our environment, so I don't really know. From my point of view, I can't answer that one.

I have used technical support but to me tech support for the EF is the same as the tech support for the rest of the NetApp environments: pretty quick, pretty easy. It's a lot more fluid.

I requested it, but I didn't make the decision.

We knew we needed to invest in the EF series mostly because we have a lot of files, we have some CDOT systems that we have in there with the 8060s right now, and we knew that we had some environments where we're looking at XtremeIO and so on. We were trying to find something comparable to it and, honestly, within our company, there wasn't a whole lot of knowledge that there were more options.

I brought it up and said, "Hey what about these guys?" And they said something like, "Oh, they don't have it." And here it is. That was what brought it in. We were using XtremeIO and I wanted to move over to this, because of cost.

XtremeIO, when it got bought out by EMC, the cost went up and the support model for EMC is the same across all products. However, now it's even more, because you have to call EMC and then EMC points you over to a third party and it's troublesome.
The advantage of XtremeIO is the GUI system, which is extremely easy. It really is. It's based off of almost like the old XIV. The XIV system from IBM was an extremely easy GUI, just slide. It's almost like using a Windows system and they kind of ported that into the XtremeIO, which made it easier for that as far as the manageability side of it. However, as far as flexibility, it didn't have a whole lot there.

Besides NetApp and XtremeIO, we brought another one in-house and it didn't make it through the first stage. We pushed it and actually crashed it. You shouldn't see a flash crash but we did.

Reliability is the most important criteria for me when selecting a vendor to work with.

I was involved in the initial setup; I helped built it out in the lab.

We normally have the vendors come in and actually do the physical build-out of the systems themselves but then, once they start doing the next part, which is the configuration, they bring us in, it's simple. It was very simple. There wasn't a whole lot for us to do.

Make sure you truly test the possible solutions one-on-one against each other and not just let the vendor tell you the answer. A lot of times, their answer is dependent on the criteria that they use to give it.

Having the option of such high-speed storage in the data center is what makes it valuable. When a request comes in a for an application that requires something on flash, I have the EF to go. I know that whatever the application is, it can't beat it.

They could improve the product’s tools. We're going to tie it into our FAS system because the options we have with the FAS system are much greater than the options we have with the EF series. Things like the FlexClones, SnapVault, SnapMirror, all of that. Some of it's available on the EF series, but we like what we have in the FAS system. If the EF were to have that, we would not need to tie it into our FAS system.

It’s a stable solution. We have not had an outage in a year and a half.

It is scalable. We're about to roll it into our other NetApp products. That, along with the FAS system, makes it scalable for us, at least.

Technical support is good; maybe not as good as others, but they're good. It’s a case of finding the right person. NetApp's a pretty large company, with a vast array of protocols and products to dive through when you're trying to troubleshoot a problem. It just depends on the person you get when you call in. Sometimes, it's the best. Sometimes, it's not so much the best. They're good. They could always be better.

We have multiple hybrid solutions, but nothing pure flash like the EF. We have been using the hybrid solutions for five years. The trigger to moving to EF was that the cost of flash has come down, and the need for flash has gone up.

For initial setup, we had somebody onsite. We were doing the flash install along with the transition from 7-mode to CDOT. We had professional services onsite to assist us with that.

What is your use case? Depending on what your needs are, I would point to maybe the All Flash FAS. If it was just a one-off, one project, throw everything at it, then I'd say definitely the EF would be the most cost-effective solution.

The most valuable feature for us is the ability to move the workload between peers. It also has a huge advantage of helping us cut down on power consumption.

Almost all of our infrastructure is on cDOT, and now we are able to have a single point of management for all our data.

It's very difficult to manage the CLI. Also, the license model needs work. If you read the label as probable, you need to pay for a license for all the features, even if you are using just one.

We've been using it for one year.

I guess at the end of this year or the beginning of the next, we will finish the migration of the whole infrastructure.

9/10 - The problem is about the failover of the interfaces for some problems. We had some issues last week because of the interface of a cluster physically remained up, but we had a report noting issues with traffic management on it.

8/10

8/10

No previous solution was used.

It was simple.

It was done in-house.

We looked at EMC and IBM, but we chose EF because of the need to take snapshots and to save it to another system.

It must be evaluated depending on the workload of the applications.

It's fast! It can perform one million IOPS.

The database queries on our old system (HP EVA) took nearly 32 hours, but on the new one in under 10.

Monitoring and management handling in enterprise environments could be improved.

We've used it for one year for our Oracle database.

We've had no problems deploying it.

We've had no problems with stability. It runs consistently.

It has scaled to our needs.

Customer support issues have normally been cleared up in one business day, so it's been really great.

Normal issues like performance problems and parts replacement are infrequent and are taken care of quickly.

We previously used HP EVA, which was slow and old. We switched from HP to NetApp because the customer got a 7-Mode FAS for a different situation and asked what he could do about the slow speed of his EVA. We recommended EF.

In two hours, it was up and running.

I implement it for our customers.

At first, we thought we could expand the 7-Mode FAS for our client, but the workload was not matching to the system, so we chose the EF.

If you need very low IOPS and throughput and an easy to install and stable solution, then this is the perfect one for you.

The most valuable feature for us is its low latency, which is the main reason we bought it. We had performance issues before, and this was a great solution to those problems.

We run our data warehouse on it and batch processing needs to happen in a certain time frame. If we can’t deliver data processing in eight hours, then we have to keep the system closed for end users at that time. Ideally, we need it done by the start of business in the US, because if we finish late, people can’t start working until after that. Our batch processing time went down from 14 hours to seven hours with this solution.

It says NetApp on the badge, but it’s not a NetApp product. NetApp acquired the product so it is missing a bit in the integration and management performance monitoring.

We've used it for one-and-a-half years for OLTP databases.

It's very stable, and we've never had any problems with stability.

We haven't had to scale it, but I'm sure it will scale.

We haven't had to use customer service once yet.

We haven't had to use technical service once yet.

We used a traditional SAN array and replaced it with flash.

The initial setup was very straightforward.

The situation we were in before limited people’s ability to do financial reporting and forecasting and analysis. Now, it’s not a problem because we are able to back everything up overnight and not cut into people’s work day.

It cost us $160,000 to implement.

• Pure Storage
• EMC XtremIO

We went with NetApp because price per performance was better.

Before investing in flash, you should be really sure that it is solving your problem. For us, it was really clear that latency was a big part of our issue, so flash was the solution.