VMware vSAN Reviews

Our primary use case for vSAN is server virtualization. We've used it to virtualize close to 500 servers which would normally have been on physical hardware. We have virtualized and consolidated it down to run on nine nodes of vSAN. That workload primarily consist of web servers running Linux or Windows Servers to support the Windows Active Directory that we have for the environment onsite.

It's improved the organization overall primarily because the storage is local on the boxes. Before we were with vSAN, we were with another iSCSI product which was a clustered product that went across the network. We had multiple instances where we would have either a network hiccup (caused by us) or a network hiccup (caused by the device). This took a whole bunch of VMs down with a lot of repercussions. It took a long time to recover. By eliminating dependency on that back-end storage, we now depend on everything that's in the VMkernel with vSAN. So, we eliminate the middleman.

We like that it is a hyperconverged solution. Everything is in a box. You got the compute, memory, and storage. So, we can scale out by adding nodes as we go and eliminate the back-end storage, whether that's a NAS or iSCSI device. 

You get the benefit of local storage, but you have the protection of shared storage.

I see room for improvement with vSAN in particularly in the reporting realm. Now, with vSAN 6.7, they're starting to include vRealize Operations components in the vSphere Client, even if you're not a vRealize Operations customer. So, that's really good. It exposes some really low-level reporting. I would like to see more of that. However, you have to be a vRealize Operations customer to obtain that. I would like to see more include of this included in the vSAN licensing.

The vSAN licensing is not an inexpensive product. It does cost more than hypervisor. I would like to see more basic reporting, or even expert reporting. I think with our licensing that we've paid our dues, and we should get the information.

Stability is working very well. vSAN is very dependent upon your network. If your network is stable, vSAN will most likely be stable. 

Our network is very stable. Therefore, we have not had issues.

We started with a three-node cluster. We are now at a nine-node cluster. We can just add nodes piecemeal as needed to add capacity. It's been very transparent. Users have never noticed when we've had to do that. So, scalability has worked real well for us.

We've been with vSAN since the early days of ESX 5.5, when it first went general availability. In those early days, we used support quite a bit. They were very good. The vSAN team that VMware has are top-notch. I think they pick the best of their support people and make them vSAN representatives. In the early days, I used them a lot. Not so much lately, because the product has gotten so much better. 

I was involved with the initial deployment of vSAN at our site. The most complex thing is you have to live and die by the vSAN HCL list. You can't put a product or a component into a vSAN node that is not on the host compatibility list, particularly the SSDs and their firmware which is specified on the HCL. You have to match that explicitly to receive good results.

I see ROI on vSAN because we have gotten out of the business of depending on the back-end NAS device or the back-end iSCSI device. We get the return on investment by decreased administrators' time, decrease exposure to network issues and stuff that would take a lot of VMs down. That's where we see our ROI.

We looked at Nutanix before we went with vSAN. For budgeting reasons, we weren't able to pursue Nutanix after a pilot.

The product is at least an eight to eight and a half out of ten. Because the feature growth that I've seen them put into the product since we've been with them since 5.5, they are innovating with each release. They're adding more features and all that adds up to a better ROI on our investment.

As we were consolidating so many servers, we had a really high consolidation ratio. We wanted to have something that was close to being local disk. However, we also needed to have redundancy so we could take a node down for maintenance or if a node would crash. All the same standard reasons of why you would want high availability.

What I look to see in a vendor is good customer support. I want to talk technical with someone. I don't want a lot of marketing PowerPoint stuff. I want to talk to people that know the product very well. Because if I start using the product, I will need that support on the back-end. I don't want to be flailing by myself in the wind. I want to have good expertise that I can call on to help.

Primary use is just for VMDK storage. We're running an all-flash array with NVME caching tier. The performance is really good, we're using SATA drives. We're about to do a complete rebuild with 12-gig SATA drives as the capacity tier, and bigger, newer, faster NVME for the caching tier.

vSAN has improved our organization by giving us yet another high-speed data store. Previously, we were using VNX that had some Nearline-SAS drives with some SSD caching on it. But the all-flash vSAN is obviously much, much faster. We also use a Pure Storage array that we just got in a few months ago.

The most valuable feature would be: You own the hardware already. Why not just throw some drives into it and have a software-defined network storage system?

I know they're working on this: better support for an all-NVME array. Better metrics.

vSAN itself is a great storage platform, but one of the issues with it is that you have to be fully locked into the VMware package to use it. We're going to be deploying 72 Kubernetes nodes, and we're not going to buy VMware licenses for 72 of them, just so they can access vSAN. That's what we're using the Pure for. Opening it up so you could have vSAN as a data store, use it as a data lake, hit it with an NFS, S3 from outside the VMware ecosystem, would be great.

Stability is okay. We do see weird things crop up every now and again. It will say that a drive gets kicked off even though it's fine, and we have to re-add it. So a few gremlins here and there, but for the most part, it's pretty good.

So far, for scalability, we've just been running it on five nodes at our primary data center, and we're building out a second data center. It's going to be running on five nodes there. We haven't really scaled it up since we built it.

I've had to use tech support once or twice. It went okay, as with any tech support.

When we started with VMware, it was a three-node package with the VSA, virtual storage appliance, which was sort of the precursor to vSAN. And it just came as a package, so we said, "Okay, great. We have our storage and our compute tied together."

I'd say vSAN, on a scale of one to 10, would be a seven or an eight now. (If I have to choose it's a) seven. But with what I've heard while I've been at VMworld, I'd say that they'll probably go up to an eight.

Our primary use case for vSAN is for our corporate cluster, and we have many different use cases using vSAN. It was a perfect solution for us. We were there for the beginning of vSAN. We created our own vSAN environment with their early installers and now we have a professional one. It's a great solution.

vSAN improved our organization by taking a whole bunch of servers that we had that were depreciated and letting us remove all of those workloads and put them on one, centralized solution, and have great storage in the back end. It's really helped us consolidate a lot of workloads that were in different silos, and now we're back to managing everything from one place.

The valuable features of vSAN are that

• you can get it up and running quickly
• you get redundancy built-in
• it's pretty much the perfect solution for a cluster.

The product can be improved in a couple of ways. One of those would be that they have a lot of hidden features, that are through the CLI, that would be great to have in the GUI, or just be more open about those features. It's something called RVC. It's a tool in the back end. It's a really great tool, but I had to find it through Reddit. So more information on stuff like that would be great. 

Also, in the user interface, giving us more features and more reporting that we can do from vSphere itself would be helpful.

Now it's great. The stability of vSAN is getting better every day. We had some hiccups in the past, but we worked through it with some great techs. They were there with us the whole way, and we got through most of our hiccups. 

There are definitely some things you need to know about vSAN going into it, like don't over-commit your storage, that we didn't know. We hit every problem you can probably hit with vSAN, but we're good. We're still up and running.

We started with three nodes, added a fourth. It was easy to do, gave us more storage, very scalable. You can just keep on growing and growing.

I was involved with the initial setup. It was fairly easy to get up and running, at first. We had some networking hiccups here and there but, overall, it took about a day to get us ready to go.

The ROI data on vSAN: I would definitely say it's my staff cutting their time by something like 90 percent. They're only dealing with one stack of servers right now. All of them are able to perform the storage tasks needed. Everyone can manage it. We don't have to wait for that one guy to come in and do what he has to do. My entire staff is trained on vSAN. We usually spend no time in it. Before, we were dealing with a lot of different solutions that took up a lot of our time, so time saved is a good reason for our ROI.

If I had a colleague in the field, what I would tell him is that vSAN is great. I would do four nodes instead of three. Make sure that you're safe. Four or five will get you right where you need to be. You won't have any problems. That would be a tip I would give: Go for four nodes. vSAN is definitely worth the money.

I would say it's a nine out of ten. It's not perfect, but it's almost there, and it's great.

We're primarily using it in a VDI environment, a four-node VDI environment. Performance is very good. We're very happy with it. Networking setup was a little bit of a challenge, but we got around that.

Reduced complexity. We don't have to worry about the physical SAN anymore. That makes it easier. The learning curve as well, when people learn vSAN, they find it very easy to manage compared to a physical SAN.

Flexibility, growth, and expansion are probably the more important features for us.

As our environment grows, the more users come on, the more VDI workstations that we need, we can easily expand either horizontally or vertically with the environment. We're very happy with that.

A bit more information on the upgrade path, upgrade availability, how to upgrade, that would be very useful.

We find the stability very good. It really reduces our overall operations.

We find the scalability very good. We've been able to upgrade very easily as users come on, as we need to create more VDI workstations. Adding the extra drives gives us the capacity we need.

We haven't needed to use technical support so far; nothing at all.

Up until about a year-and-a-half ago, we were using physical SANs. Space is a problem in our environments that we deploy, so we knew we had to get rid of the physical SAN and go toward the more virtual environment. The number of nodes we deploy, we need them. By integrating the vSAN, we're able to get the space requirements we need.

I was involved in the initial setup. In fact, I was involved with the selection of vSAN compared to other products, as well as physical SANs, and I was involved in some of the design and configuration.

It was fairly straightforward, actually. After we got around the networking issues, we found that the vSAN setup was very good.

In terms of return on investment, we don't have any kind of requirement there.

We considered EMC as well. We considered HPE LeftHand, which we had used in the past, so we were familiar with the virtualized SAN. We like the vSAN a lot.

The advice I would give is to properly analyze your host infrastructure. Make sure that your network cards are sufficient for the environment you're trying to deploy in, whether it be all-flash. There are already some Ready Nodes available. Go with the Ready Nodes when it comes to vSAN. Don't try and buy your own parts - something we looked at originally that we scrapped. That would be my main advice. Go with Ready Nodes when it comes to virtual SAN.

In terms of improving the product, we're very familiar with the new features in 6.7, which we're going to be upgrading to. Data encryption, we would like to deploy, as well as compression and deduplication. Those features are already available in the new version. We just have to take the time to deploy them.

Out of ten, I'd give it an eight. We're very happy with the product. To bring it to a ten we'd rather not upgrade as often. Right now, we're at 6.2 and that wasn't long ago. They're already going to 6.8 now. We'd like to have a little bit of a normalization period before we get to the next product. I understand it's a focus for VMware. We're very happy they're focusing on it.

We use it for our DMZ and any test environments that we put into our industry.

It's performing pretty well. We have no issues with vSAN at all.

It has improved our organization in a way of scaling it. 

• Cost was big for us.
• Speed
• Scalability

We can scale it very easily for a test environment. We were able to segment our DMZ so it wasn't connected to anything, which we really liked.

One thing in vSAN that I would like to improve is using vSAN as a repository for files or other things. For example, with Horizon, maybe we can save profiles with UEM on there. That would be a good feature that I would like.

The stability has been great with vSAN. We have not yet seen downtime.

We scale it with our test environment. We are looking to do it with Horizon. We are able to scale it to see how many VMs that we can host and how long it will take us to add new hosts, if needed.

Technical support has been very good. They respond pretty fast, especially if we have a critical issue. Their responses have been great.

vSAN is one of the easiest implementations of any VMware product. It's almost like click it to enable it, then you're almost done. So, vSAN is very easy to set up.

We did consider other hyperconverged solutions. It usually came down to price. vSan was the most cost effective thing. That's why we went with it. Also, we didn't have to get a connected array. We can put it in small places, remote sites, etc.

Nutanix, Cisco HyperFlex Edge, and VxRail were on our shortlist.

I would rate the solution an eight out of ten. To make it a ten, it needs to be able to scale the amount of data that we can hold so we can put bigger, more data-intensive apps on it.

My advice to a person looking at vSAN is get your hands dirty in the labs. Show how easy it is to set up, because it's not very complicated. It's an easy solution that you can implement at your company.

Most important criteria when selecting a vendor: Since we're a hospital, we have multiple hospitals in the area. We look at local site resiliency, so we're looking to see if we can put it in each of our hospitals.

We recently adopted vSAN. We adopted VDI for our desktop solution about ten years ago and we have a single KPI for delivery which is clinical data accessed in five seconds.

Throughout the last decade, as new back-end technologies have come to market, we have always been investing in the hosting end of VDI. Five years ago, we went to an all-flash array, and two years ago, we went to the vSAN hyperconverged. 

When we went to vSAN, at that point in time, we doubled the density of our desktops per host and, for the first time ever, I could demonstrate a significantly lower TCO for a VDI desktop versus a rich or fat client.

For my organization, the most valuable features of vSAN are as follows:

1. The most important to me, in my role, is cost. In the renewal cycle for storage, it was about a 40 percent saving compared to going to an all-flash array, which is what we first looked at doing.
2. Performance: our clinical data access in five seconds; we need to do everything we can to retain that metric.
3. I was really pleasantly surprised during the data migration across to vSAN, that it happened almost instantly. Whereas, in the past, migrating from array to array was an arduous and fraught process.

Room for improvement could be in the planning stage of going to hyperconverged. And this is a big ask: some modeling tools or guidance on how to work out the optimal TCO. For example, core size - the amount of RAM that you're running - versus the licensing cost you're up for with, say, Mircrosoft data center, versus the number of hosts you're going to run and have to license for the vSAN. It's quite a complex equation and it's really difficult to work out, in advance of implementing the solution, that you've got it right. That creates some uncertainty around the total cost of ownership.

Stability on the vSAN has been 100 percent. As part of the implementation process, the VMware customer success team for vSAN assisted us. We actually retrofitted hard disk into our own existing hosts and they went through a process of review and remediation to get all the "green ticks". We went through that process in advance of putting it into production for our data center, which we did this year. So, there have been absolutely no problems from that perspective.

When talking about scalability, the real value is that, for the first time, I can just build it out one host a time. Over the years, I'm sure everyone has experienced hitting the wall on their array where it's too old or the technology has changed, and they're up for a large sum of money in one hit. The actual, repeatable, non-quantity of the cost to increase the storage, is very valuable.

On a scale of one to ten, I am giving it a nine. It's probably because I can't bring myself to give a ten for anything, in case it could be improved.

We use it for all our virtual desktop storage.

It's definitely cheaper to buy it piece by piece, instead of an entire shelf at a time.

• It's easy to scale.
• It's easy to predict IOPS needs.
• You can design for low latency using all-flash.
• The whole hyperconverged notion is pretty neat.

Also, for setting up new clusters for VDI quickly, it's nice. You don't have to wait on an order for a storage vendor to ship you a system and help you configure it, you do it all yourself. It's kind of convenient that way. And the sizing guides are pretty straightforward.

I would like to see better performance graphs, maybe something that you can export outside to a different console, and maybe a little bit longer time period. The 18-hour maximum, or 24-hour maximum, is kind of short.

Also, the hardware compatibility limitations are a little frustrating sometimes, but as everybody's starting to adopt vSAN more, you get more options for hardware.

It's stable. We haven't had any major issues.

Scalability is easy. You just buy a node and go.

The vSAN technical support guys are great.

We chose it because of cost considerations. We already had an enterprise agreement with VMware, so vSAN licensing was included.

There was a small learning curve, but it's pretty straightforward once you understand the basics of how everything works.

We did evaluate other vendors initially but this was our second hyperconverged solution. We went with it because of the cost.

Do your homework. Make sure you know what kind of IOPS and latency requirements you need to meet. Picking hardware is not hard anymore. Everybody has an HCL. vSAN has a great list. Just pick what you want and go, it's not that hard.

I rate it at eight out of 10 because nothing is perfect. I'm hard to please. I'm not saying there are growing pains, but vSAN was still new at the time. They didn't have dedupe and compression yet. The performance was pretty good. Most of it was hybrid in the beginning, but now with all-flash, it's speedy, when it needs to be. It's a young product and nobody gets a 10 out of the gate.

We use it for VDI.

It's supposed to provide low-cost for storage arrays to do VDI. We're on the fence with it. We're still looking at other solutions. We're not sold on it.

It has provided some value when it's working. Instead of hitting our production SAN array, it has its own array, storage-wise. It keeps workload off production.

It could be more robust. The latency is also an issue for us, and the reliability. I would like it to be faster and a little more flexible.

On a scale of one to ten I would give the stability a six.

Scalability should be pretty good, but we're not getting the performance we want out of it right now, so we're not going to scale it unless something changes.

The initial setup is pretty straightforward.

We have seen value in it but, since it's not performing the way we think it should. We're probably not going to move forward with it.

We went with it because of the cost. It's definitely cheaper than buying a storage array.