NetApp AFF (All Flash FAS) Review
Integrates seamlessly with what we're used to for FAS while getting the raw performance of flash

How has it helped my organization?

One example is we're moving a legacy application over. I'm actually in the middle of a project for that right now, where it's four Windows servers each with eight terabytes, that our actuary department uses for data analytics. With the efficiencies on the AFF, that eight terabytes has gone down to about two and a quarter of actual capacity used. So we're going to save a lot of space there, in addition to letting them run more simulations and get more simulations done more quickly because of the storage being so much faster than what they're on now.

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

What is most valuable?

Some of the best things about AFF are that it integrates seamlessly with what we're used to for FAS as well. We can use the same ecosystem, OnCommand Unified Manager, but get the performance, the raw performance of flash. It's great that way.

I think that's the most important thing, the integration with the existing features that we already have and existing management systems. Among those features are the ability to do SnapMirror or SnapVault for data resiliency and backup. The other features are the data efficiencies, compaction and inline dedup compression, that let us use it more efficiently too. Those are huge on the list.

What needs improvement?

Looking at the road map that's out there, I think they're heading in the right direction. Additional performance, additional data efficiencies, that's what everybody wants right now.

And then the integrations that I'm really excited about - and part of the reason I'm here at the NetApp Insight 2017 conference - is to look at the integrations with AFF and things like StorageGrid Webscale. So you're getting even more efficiency out of the platform and offloading cold blocks that you don't need right away.

What do I think about the stability of the solution?

We haven't had any issues, even going back to the longer experience I have with the FAS platform. They're typically few and far between, especially compared to some of the other vendors we've worked with. When we do uncover an issue, we typically get escalated to the right teams and get it worked out.

What do I think about the scalability of the solution?

It's really good. There are some that things that could be done better there, like NetApp is doing; it's other products like Webscale and SolidFire. As long as you're aware of the design considerations, it's very, very easy. Shelves go in like a snap. As long as you make sure you have the proper compute to go with it, you're good to go.

We're not really having scalability issues, it's just you have to make sure that you're not exceeding the capacity of your heads when you're expanding your logical storage out, that's all.

It has caused problems for my company in the past, but I think that was the result of not having storage administrators with a high level of proficiency and knowledge of NetApp. They made some very poor sizing decisions, but you can't blame the vendor for that. It's more of the admins' fault for not specking them out properly.

How is customer service and technical support?

For the AFFs, I don't know if we've had to specifically leverage NetApp support yet. I don't think we've had an issue major enough that we've had to reach out. That's been more on the FAS side.

Support has generally been pretty good. Occasionally there are struggles getting to the right people but, once you do, they know what they're talking about.

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

Which solutions did we use previously?

Yes and no. We're in the process of retiring some old storage frames, old Hitachi frames actually. I believe it's just disk-based. There are actually three different Hitachi frames and they're different. One is all flash, one is hybrid, and the other one is purely disk-based. So there's a mix. We have another all-flash platform that we could move workload to, but the NetApp fit the workload a lot better for this in my opinion. So it made sense.

The original intent was actually to extend our NAS - we primarily use NetApp for NAS and a lot of our environment. But we've pitched the AFF that we just installed, the A700, primarily as a SAN platform. So we're really trying to leverage more towards that now.

It will eventually be used for both block and file storage. It was originally slated for file usage NAS, but we're leveraging it more for block.

I had worked with NetApp as block storage in the past, and I always had a high opinion of it. I think NetApp is the best in the industry at providing a unified platform for file and block. Hands down.

We don't get too deeply involved in the cost analysis, but management and engineering rely heavily on the input from myself and my co-worker on the storage team, for these kinds of decisions, on a technical level.

How was the initial setup?

We had Pro Services, but we were heavily involved.

For someone who is experience with any NetApp platform it's very, very straightforward, very similar to anything else that you would do. Obviously there are some specific guides, specific to AFF. You want to make sure you're following those best practices, but other than that it's a cinch. It's something that I could have done on my own without Professional Services, that's how easy it was.

Which other solutions did I evaluate?

We have storage frames from most of the large vendors, so EMC would have been on the table, IBM would have been on the table, Hitachi. And really with the ecosystem that NetApp has built up around it, it just makes the most sense from a management perspective for sure. And the performance and value for money is there as well. It's a tough combo to beat.

What other advice do I have?

We have a 8080 EX HA pair, an 8040 HA pair, and an A700 all in the same cluster. That's our production cluster. We also run an AFF8040 for non-production and then a couple of other FAS heads: two HA pairs, 8040s for DR. So we've got some NetApp spread around.

Based on our experience with AFF, we are definitely more likely to consider NetApp for mission critical storage systems in the future because it's the same quality and the same value for money as we have always come to expect from them.

This is the direction the industry is going. My personal opinion is that SaaS 15,10k is going to be dead, completely within the next three to five years. Everything is going to be flash for performance and cheap and deep SATA, probably object storage for archival. I just think this purchase puts us better in alignment with where the industry is headed as a whole, it's more future proof.

When it comes to the most important criteria when selecting a vendor to work with I think what's important is performance, value for money and, in addition to that, having support that's easy to work with, that can get you the answers quickly when you need them. That is the other big thing.

I give it a nine out of 10 because there's always room for improvement. I don't think anything is perfect in IT, but it's pretty darn good. It's really pretty impressive technology when you get it running.

What would make it a 10 goes back to what we talked about above, with the additional integrations and single panes of glass and getting a whole functional flow; what NetApp keeps pitching on the roadmap as the "Data Fabric," getting a single pane of glass for everything in your infrastructure and tying it all together.

Advice as far as choosing a solution? Everybody's requirements are different, but if they don't have NetApp at the top of the list as candidates, they're doing something wrong.

Disclosure: I am a real user, and this review is based on my own experience and opinions.

Add a Comment

Why do you like it?

Sign Up with Email