What is our primary use case?
We have a variety of things. We have, like, user home drives. We have just a lot of human-managed unstructured data. We also have a data store for an AWS system, which is an automated workflow system. We store millions, if not actually billions of images. We've got stuff all over the place, but we also host storage for Linux and Unix servers that have NFS mounts for multiple applications. So it's a hodgepodge of a lot of different things.
How has it helped my organization?
A lot of it is just the ability to dedupe the data and compress the data because we had some users that are trying to move out into the cloud, into AWS and Azure, and they're starting to realize that their data footprint out there, say in S3, for example, is much greater because it doesn't have the native ability to compress and dedupe data like NetApp does. NetApp is very efficient. For the most part, we love NetApp.
What is most valuable?
The SnapMirror is a good tool because, as long as you're going NetApp to NetApp, it's ultimately the fastest way to move data. We replicate everything to another site for disaster recovery.
We like the SnapMirror technology or Snapshot technology that allows us to recover directly off the NAS rather than backing it up to a secondary device.
Just the ease of management, and it's very highly scalable. We've got one cluster that has about one and a quarter petabytes of data, and then we replicate that to an off-site. So altogether, we've got over three petabytes of data. And if it wasn't compressed and deduped like it is, we probably have more like eight or nine petabytes of data. So it does a really good job with that.
AFF is obviously the faster model because our AFFs are all flash, and they're a lot faster.
What needs improvement?
There is an area of improvement in support. So when we do have a problem, we always want it resolved right away, and sometimes it takes a few days to resolve things.
So, the response time could be a bit faster.
For how long have I used the solution?
We have a long-running history with the NetApp FAS. I started in 2006.
What do I think about the stability of the solution?
There are rarely any issues related to lagging, crashing, or downtime. Most of the time, there might be performance issues, but more often than not, these are related to factors external to our systems, such as network issues.
From a pure reliability standpoint, these systems can remain operational for extended periods. I recall instances where they've been up for about a year and a half without requiring a reboot.
What do I think about the scalability of the solution?
We could potentially scale up to 52 nodes. However, our most extensive deployment right now consists of 6 nodes.
How are customer service and support?
For the most part, the customer support is pretty good. When compared to other vendors, we have frustrations now and again. However, that's just because a lot of their support model has moved overseas, and so it's a different experience.
Thecustomer service and support are decent, but there is room for improvement.
How would you rate customer service and support?
Which solution did I use previously and why did I switch?
At the filer level, we began with NetApp and FAS. When I joined in 2006, we were already using FAS. Around 2011, we started incorporating the AFS or all-flash systems.
How was the initial setup?
As long as you're following a documented process, the initial setup is pretty straightforward, but there are a lot of steps. There are multiple things you have to do – setting up SnapMirror, configuring schedules, setting up snapshots and their schedules, and establishing all the mirroring required for disaster recovery.
So there's a significant number of steps, but I wouldn't say it's overly complicated.
We did it where we set our own up. And then lately, because our team has shrunk, but our storage hasn't, we usually have our third-party vendor come in.
As part of the procurement process, we select to have them do professional services that come up and do our installation. Or we'll do a hybrid, like the last one we just did, where they did all the on-site work and then got to a point where it was turned over to us to administer, like all the fine-tuning.
We have a hybrid setup. Some clusters are in our data center, while others are in both Azure and AWS. We've also deployed the newer FSx systems in AWS due to cost considerations. Our internal customers are looking for more cheaper storage, and FSx fits that bill better than CVO.
What about the implementation team?
We deploy systems on an as-needed basis. Our big clusters, ranging from about a petabyte to a petabyte and a half, can take a couple of months from receiving the equipment to completing the process. Most of that time is spent replicating data. However, if we're adding to an existing cluster, that's a more seamless process and can be done in around three weeks.
Our team handled the deployment. We have a group of four individuals responsible for handling these deployments.
There's occasional maintenance. An interesting aspect is that our Windows team conducts monthly patching. As for us, we mostly focus on code releases.
NetApp has been providing around two code releases annually for many years. We tend to do a major code release once a year, which is quite significant given the numerous customers on our clusters. The actual upgrade process, handled by us, takes around three and a half hours for the largest setup with multiple virtuals.
What's my experience with pricing, setup cost, and licensing?
The pricing ultimately depends on the capacity and the controller. It varies. We have systems that cost around 50,000 for the lease term, while others range into the million-dollar territory. It all hinges on the capacity and the controller specifications you opt for.
Which other solutions did I evaluate?
I am a NetApp administrator. We are looking for a data classification and data governance tool, and we've already got some in-house, but we're just looking at multiple products. So I'm literally just doing research.
However, we do have a vendor, and they are going to put together a demo for us.
We support AFF NetApp. So AFF and SaaS systems. And we also have CIFS, and we do have a little bit of FAS and AWS. So we're a NetApp shop.
What other advice do I have?
Make sure you undergo some training before diving in, as it's a lot different experience than just managing a Windows server. Windows servers are pretty intuitive and easy to manage.
NetApp systems are somewhat more complex, though not excessively so. If you're transitioning from a Windows server environment to NetApp, get training or education; otherwise, you might struggle with this solution.
Overall, I would rate the solution a nine out of ten. When other service providers attempt to sell us products during proof-of-concept trials, they usually don't measure up.
Disclosure: My company does not have a business relationship with this vendor other than being a customer.