NetApp AFF - All Flash Reviews

• The speed
• Performance
• The storage efficiency is very good.
• We use the SnapMirror for disaster recovery, though not for tiering.

By moving everything to the All Flash Array, our outage times have gone dramatically down, if not disappeared completely, for the most part.

We are more likely to consider NetApp for our mission critical storage systems based on our experiences with AFF. We are actually moving all of our production data onto our AFF system right now as it's been extremely fast and stable.

More reporting on a granular level within system command.

I have personally used NetApp for 15 years now. A long time.

It's extremely stable. We've never had an issue with it, even through multiple OS upgrades.

It's excellent. We added a shelf three weeks ago and it took less than 10 minutes.

We haven't used technical support yet.

However, I would recommend if someone is researching NetApp and similar solutions that they take a look at the support offered by other companies and look at what Netapp offers as well.

Our existing solutions, which were Netapp, were basically just overrun.

I was involved in the initial setup and it was straightforward.

We looked at not only Netapp, but we looked at EMC which was the big one. Then we started looking at some cloud providers, but we actually moved away from that.

We had a very high impression of NetApp as a vendor of high-performance sound storage before purchasing AFF, and an extremely high impression of them afterwards.

The primary use case for our All Flash system is VMware. It's NFS-based, therefore it's NAS-based.

Most important criteria when selecting a vendor:

Typically, with vendor selection, it's going to be more about the support after. Most of the features across the vendors that I've talked to are pretty much on par with everybody else.

Clustered Data ONTAP

Low latency

My company uses mostly NetApp products, so I have existing knowledge of using their products.

We're still testing it, but I expect it to provide us a lot more speed because of low latency and clustered Data ONTAP.

They still have to reduce in price when compared to their competitors. Also, the current version has some problems with global deduplication.

It's not yet in full production, but I have been testing the product for few month now with VMware.

I've had no issues deploying it in our test environment.

We do not expect any issues, and have had none so far.

It has problems with deduplication when done globally.

Their customer service is one of the best, and I am a demanding customer.

• Tier 1 – room for improvement as they hold onto tickets for too long
• Tier 2– much more serious

From my experience with NetApp products, initial setup is going to be nice and easy. We are very techy, so it was easy for us.

The majority of work was based on existing knowledge, but we also got help from the vendor.

We considered Kaminario and XtremIO. We chose NetApp in order to utilize current resources.

It has better adaptation than pure flash solutions such as XtremIO. It’s important to learn the weak spots of the suppliers in the market, and I can say that I have great expectations for the migration of the flash array to disc via cluster.

• Clustered Data ONTAP
• Data recovery
• Protection

Our company has a web application for post-lab test results, and with flash, the application is much faster than it used to be. The customers can see results much faster and can then send new requests for labs faster.

At the moment it's perfect to me because I haven’t used it that long.

We've used it for about a month with different applications for laboratory software.

We had no issues with deployment.

So far the stability is excellent

It scales to our needs.

The customer service is pretty good.

The technical support is pretty good.

We used NetApp FAS previously and switched to the flash version.

The initial setup was generally straightforward, though there were a couple areas that could be streamlined or made a little more clear in terms of which direction we should go.

We used a vendor team.

It's too early for us to tell.

No other options were evaluated.

It’s a pretty good solution; go for it.

It’s really reliable and fast. My customers have several other NetApp systems, but here they really needed the low latency.

Also, clustered Data ONTAP is valuable.

It sped up our data queries and we can work more efficiently. These queries are now done in seconds rather than minutes, affecting everything we do on a day-to-day basis.

I haven’t encountered anything the customer needs that this solution can’t do.

We've used the 8080 for eight months and the 8060 for two months. They're used mainly for VMware products and SQL databases.

I haven’t had any stability problems yet with our current customers.

It scales to our needs.

8/10

We haven’t needed to escalate to NetApp’s tech support.

I went from working with NetApp FAS to AFF, so I stayed in the family. The customer was satisfied with their FAS system but needed lower latency, so it was a clear choice.

The initial setup was very straightforward. When you've worked with NetApp for a while, it is very easy to setup their new systems. If you are new to the system, then there is a learning curve, but NetApp provides great documentation.

I implement it for our customers.

It’s a great choice, and you are on the right path using this product.

We do storage across the United States.

We have SQL clusters across the United States. It has sped up our IOPS and made it a lot easier for users.

• Uptime
• Ease of use

I would like them to roll in global monitoring instead of having to buy another product for it. If it was built into the solution, that would be awesome.

We haven't had any issues, so far.

We are scaling up to the new solution. We haven't had a lot of scalability yet. We are looking forward to what it can do.

Our technical support experience hasn't been very good. However, we are hoping with our new contract that it will be a lot better.

We were using HPE EVAs, which are very clunky and old, so we moved over to NetApp.

We were just bought out by another company who has been using Dell EMC. They're not happy with that solution, so we brought them into NetApp.

The initial setup was a little complex, because we weren't very knowledgeable in the NetApp at the time. We were using a third-party, and they didn't have a lot of technical individuals, so it took a while to get it out.

We used a reseller, EVOLTECH. It has been okay so far. There are not a lot of technical individuals with their group.

From an application standpoint, we have seen a lot of return investment on the speeds and responsiveness of the actual storage.

NetApp and Pure Storage were on our final shortlist. NetApp just came in with a better price point that my VPs and CIO couldn't refuse.

Do your research. There are a lot of different storage vendors who have a lot things which are good. Pick the one that you feel is best for you.

• Improves performance
• reduces CPU usage
• Efficient use of RAM

• Price/performance
• High availability

Stability could be improved.

No issues with scalability.

In the first years it was great, after that it has become worse.

NetApp is getting too expensive.

HPE 3PAR.

At this point, performance is the most valuable feature. We're just putting it into production, on a pretty heavy performance-intensive workload. So far, its performed exactly how we wanted it to. Performance is the key on that particular device.

It's offloaded workload that was compromising other workloads because of performance degradation. It's enabled us to take that and isolate it; give it the performance it needs, saving other applications’ performance as well.

We don't have it running ONTAP 9 yet. Upgrading the OS to ONTAP 9 will definitely give us some advantages. From what I saw at a recent NetApp Insight conference, about how ONTAP 9 looks and feels, there are things to look at and learn how to use that, in performance monitoring tools as well. We still had some learning to do about what's available. We're using rudimentary performance monitoring. As far as that goes, the old tools are giving us what we want, but we're looking forward to upgrading to be able to take advantage of better tools.

We are especially looking for better performance monitoring. We want to be able to truly see what the load is doing at any given point in time, and especially if the user wants to know, “We're going to load this up. We want to see what effect it has on it.” We want to be able to give them real-time numbers.

Right now, that’s not easy to do. We can't get to the detailed level that we want to. We believe that that's available going forward.

We've only had it in production a short time. We've had it a total of about six months.

In the short time we’ve had it in production, six months, we haven’t had any stability issues.

We haven't had to scale it out.

We've had to call technical support only because the performance monitoring on it has given us some skewed numbers. Getting back to us on that was a little bit slow, to get us the answer that we really needed to see, but we got the answer that we needed. All is good now.

I wasn’t involved in the decision process to invest in this particular All Flash FAS, but I've been in many, many discussions about going to that technology. I'm part of our team to say, "This is what we think we're going to need based on what we've seen. This could be the right tool for the job." In general, with decisions like this, there’s no one person making the decision.

We were previously running on a different vendor platform. We had that device saturated, and there was nowhere to go with it. The scalability was non-existent. It was disk. This was a good opportunity for us to move into this flash environment with this particular workload because of the performance.

At this point, there really wasn't another player that was going to offer us familiarity with NetApp, for one thing, and what we needed.

The most important criteria for me when selecting a vendor to work with are the ease of administration of an appliance; reliability of an appliance; and being able to adequately monitor what's going on with the appliance (which ties in with the administration of it). Support’s got to be on it, especially if it's in production. It's like, “We need help; we need it now.” The vendor has to be there.

Those are probably the three most important criteria. Price comes in there, but you pay a premium for those particular things. If the price point is right and those things are all right, then you've got a great thing going on.

Flash right now is just a hot ticket. If you've got performance-intensive workloads, and because the NetApp suite of tools that can come along with it, then, yes, I would recommend to colleagues that they take a look at it.

It's still pretty new to us, but what we expect it to do, it's doing. As we get more familiar with it, and if we see that we can scale it out and add more to it, I think I would be able to rate it higher pretty easily.

The most valuable thing I have seen since we've got it is that predictable performance has stayed below a millisecond, which was not the experience we've had with spinning disk. So, I was looking forward to that coming in and giving my customers predictable performance, and it’s proven to be doing that for us.

Learn about the benefits of NVMe, NVME-oF and SCM. Read New Frontiers in Solid-State Storage.

We're having a hard time deciding what goes on flash and what doesn't now. When we're doing replication, where you have an all-flash array and we're replicating between sites, we want this flash but we want to have SATA for replication, as well, for a target. So, we're having a hard time deciding, should we go FAS or should we go all flash?

While at the recent Insight conference, I talked to some of the more senior technical guys. They were able to give me the difference in impact on performance from a FAS running SSD and an all flash running SSD. There's not that big of a gap. And so, that gave me more confidence that we could go hybrid if we need to on our smaller sites, and then still get the replication done on low cost and not lose the big performance that we got out of flash.

It has been very stable, just like the other products that we've had from them in the past.

We bought small and hoped that the efficiencies would bring in what we need, and it did. But with everything going on in our environment, we actually increased it so that we can have a little more capacity. Right now, it's probably 2% utilized, which is completely different than a spinning disk, which is 70% utilized. So, the scalability's just easy to do; it's incredible.

Support, I think could use a little bit of help. We can't seem to get to the backend guys fast enough. We've had conversations with them about that. So, we would love to see some of that going on and get better support quicker, to the right guy.

Learn about the benefits of NVMe, NVME-oF and SCM. Read New Frontiers in Solid-State Storage.

Initial setup was very straightforward, with the new optimized arrays that you can purchase and they come in with a 10-minute setup. That did take away a lot of the steps that we used to do before. So, it did come in, we were able to just plug it in and in 10 minutes have it up and running.

We were already a NetApp shop, so for us this was just adding it to the cluster. And it was time for us to do that with a hardware refresh, so we really didn't compare to others.

The most important criteria when I’m looking for a vendor are stability and availability. Cost is always thrown in there, but it's not the first one. And then support is becoming more and more important to us; being able to get to the right person at the right time.

From the All Flash, from being a NetApp customer for quite a while, having all protocols in one box is very powerful. And so, I would say, that would be a great thing to consider when you're considering the all flash array is, most of the all flash arrays out in the market today are block. They do have the file protocol, they're leading in the industry with it. And we've switched over to the file protocols quite some time ago. And we're seeing much more savings in operational costs because of the file. We take out the zoning and all of the block stuff that comes with it, and we're being very successful with file and we've reduced our operational costs significantly because of it.

I'm very happy with it and the low latency has been good. It's met the mark.