What is our primary use case?
We use it for virtualization related to development. We have two entities in our company. One is corporate, a subcontractor for NASA. And the other one is an electronic timesheet system. For the corporate side, it's mainly a file server. And we use StarWind HCA for development of the electronic timesheet system. It provides us VMs and tools.
How has it helped my organization?
We can do updates without any problems. We can move all my VMs to one host and do updates on the other host. We can bring it down, move everything over to the other host, and then update the other host and bring it down.
In terms of redundancy, with my last solution, if we had two VMware hosts and one host went down, everything would transfer over to the other host. StarWind HCA is the same concept except that we don't have the single point of failure of the storage array anymore. It's all in the hosts. We don't have to worry about the storage going down. It used to be that if the storage array went down, we were dead in the water with both hosts.
Our only real choice, other than StarWind, was to buy a Dell EMC Compellent which would have been double the cost and would still be just one Compellent. So if we wanted redundancy, we would have had to put together a solution that would triple or quadruple the cost. StarWind saved us a considerable amount of money.
What is most valuable?
The most valuable features of the solution are the redundancy and its cost. I used to have a SAN, a Dell EMC EqualLogic. Unfortunately, it was what they call an "inverted pyramid of doom." It was two or three hosts, two switches, and one storage array at the very bottom. But the SAN, the storage array at the very bottom, is a single point of failure, and many people, including me in the past, don't take that into consideration.
The SAN was working for us, but I thought about the fact that it is a single point of failure. Anything could, possibly, take it out, even though it might have redundant hardware inside it: controllers, power, hard drives. The entire unit itself is a single point of failure. If updates were required to an EqualLogic, we would have to take down everything, just to be on the safe side. We'd have to shut down all the VMs. And those updates could always mess up the entire unit and, then, it's a single point of failure and all your infrastructure and VMs are down.
In terms of cost, a storage array is more expensive. It was time to renew our storage array. It was end-of-support, end-of-life, and the EqualLogic line is supposedly being phased out. The next in line is Dell EMC Compellent and we would have had to upgrade to that. It is highly expensive. For half the cost of Compellent, I got two hosts, more storage, and redundancy.
StarWind HCA also has a much better footprint because with a full-blown SAN you have one storage array, or in some cases two, as well as two switches and two or three hosts. Those two hosts are usually 2U each, and the storage array is 2U, and the switches are usually 1U each. We were able to shrink it all down to two hosts that contain all the storage, the switches or the all the storage networking, and the host or the compute/CPU power. In total, the HCA is just two hosts and they're both 2U. So our footprint was reduced to just 4U.
What needs improvement?
One area for improvement of the solution is that I had to get the HCAs with Windows Server installed to install the StarWind SAN software on, which I really didn't want because of the extra maintenance or overhead, as well as viruses, etc. It's going to take time for them to get their Linux implementation to that point. They already have Linux but it's not as mature and they don't really support it on HCAs. They have it for individuals who want to use it on their servers, but not on HCAs.
With Windows, there's always that fear that, if you add any software to it, if you need to configure monitoring software or the like, DLL conflicts and blue screens can result. Similarly, if you use Windows Update, you can get blue screens. Or, there have been times where an antivirus company has made a mistake regarding its virus definitions and it took down the server. The antivirus blocked or deleted a legitimate OS file that it thought was a virus. So I don't run antivirus on the Windows Servers VMs that run the StarWind SAN software. At the same time, I've had to configure Windows Firewall to block everything and only allow any kind of traffic going to the server. The only thing I allow is just Remote Desktop so I can manage it. But even Remote Desktop, in the recent months, has had exploits. I keep on having to do Windows Updates.
I prefer Linux because it's not as targeted. Don't get me wrong; it is targeted for viruses and all, but not like Windows Server.
For how long have I used the solution?
We've been using it for a few months.
What do I think about the stability of the solution?
The solution has been stable so far..
How are customer service and technical support?
We are using the solution's ProActive Premium Support but it hasn't really reduced our monitoring efforts since we've only been using it for two or three months. We haven't had any issues come up where we've had to use it. I still do all the monitoring of my VMs and the hardware, the HCAs. However, in regard to the support itself, I do like that it's all-in-one. If I need support I call one vendor and they take care of everything. They call Dell EMC, they call VMware and, of course, they take care of the StarWind software. So it is nice.
Plus, each person I've talked to — and I've talked to multiple people there — has been very knowledgeable. I didn't get the sense that any of them were new or learning or that they didn't know what they were talking about. All of them are very knowledgeable and friendly.
How was the initial setup?
I wouldn't say the initial setup was completely straightforward but it's not too complex. I did have a lot of calls with support to help me get it up and running, but I did the majority of the cabling and some of the configuration of the VMs. They took care of many other things that I would not have known to do, but it wasn't too bad.
The deployment took about a month. I had other things I had to do; I'm always doing a lot of things. It probably took longer than it could have taken.
The implementation strategy was that I have all iSCSI. Our previous SAN had iSCSI with RJ45 switches. With the help of StarWind and Dell EMC, I was able to tie in and connect the HCAs to my SAN and see the data stores on the SAN from the HCAs. When the time came, I was able to migrate everything. I placed all the VMware hosts into one vCenter but two different clusters. I was able to simply vMotion them. Once I got the HCAs up and running, configured and set up, I was able to vMotion all the VMs from my old storage array to these HCAs.
What about the implementation team?
Overall, I did like the hardware installation and the cabling and they helped me configure the StarWind software. It was about half and half.
They were top-notch and professional. They know their stuff. I was always able to get them online when needed. Their support was very good.
What's my experience with pricing, setup cost, and licensing?
Regarding licensing fees, the caveat is that with the Windows-based OS, we have to pay for that licensing for both hosts. That's is another reason I wanted Linux. As for VMware, we already had VMware licenses, so we just took those from our old hosts and applied them to these hosts.
There is also a cost for the ProActive Premium Support and, on top of that, is support for the Dell EMC hardware itself. We got four-hour, mission-critical, which is what we have on everything else.
Because of the absolute redundancy of the two HCA hosts, which they say can tolerate a failure of one host plus one drive, you might be able to save a little bit of money by bumping down the support of the servers and not need four-hour, mission-critical support. You could bump it down and wait for parts to arrive the next day instead of four hours.
Which other solutions did I evaluate?
I looked at Dell EMC storage. The only option was Compellent, which was highly expensive. I looked at Nutanix which was still highly expensive.
I also looked at StorPool; I liked the idea behind it, but I didn't like their implementation. It's roughly the same concept but requires more hardware. They take a bunch of servers that are not purely storage servers but which have the compute and memory. It's a rack mount server with all the storage inside and they aggregate the storage.
StarWind was all-in-one and consolidated on two servers. StorPool would have been three servers just for the storage. I would have had to buy two more new hosts to be the compute.
What other advice do I have?
I love StarWind HCA because of the cost and the redundancy. I love the service, the support. Across the board, it was the best choice. I love the HCA because it's all-in-one and everything is pre-configured. I could have bought my own servers but it would have taken longer to bring up the environment. It would have been less expensive, but StarWind's hardware and software support and the compatibility of all the hardware components add a little bit more to the reliability of the system. That's why I went with the HCA instead of doing it myself. I certainly could have done it myself if I had more time. But, as a small business with one or two people managing all the IT, it was the best choice.
We have two environments, one at the office and one at the data center. This implementation was a trial of sorts, but looking to the future I'm going to implement this for our data center, where we have a standard SAN like we did before this HCA implementation.
The solution has not improved our system performance. There were some things that we couldn't foresee or we didn't test, like restoring databases. It's a little bit slower there. That's more a failure on our part, not having tested it out, rather than StarWind's failure.
We have a hybrid HCA as far as our drives go. Some are flash drives and others are just regular spindle drives. The solution is supposed to move things into the SSDs and then give the appropriate power, from what I remember them telling me. But in one particular case, one of the developers, who is also a database admin, was restoring a file and he said it took way longer than usual. That was one thing we couldn't assess during our assessment of what kind of drives we needed. In this case, we probably would have done better having all flash drives. It might have been overkill — it depends on what you need. But we should have made it all flash drives and we probably wouldn't have had any problems. Again, that's not anything on StarWind's part.
Everything else, performance-wise for all the other VMs that we have that are not as intensive as a database, it works just fine. We have no complaints about the performance in terms of using it as a file server or for web-based development utilities.
We're a small company. We have two entities that these HCAs provide service to. We have about 30 to 40 employees. Of them, 10 or more are on the corporate/sub-contracting side. The rest work on our electronic timesheet system, whether they're in development or technical support. In terms of deployment and maintenance of StarWind, it is just me.
StarWind gives you choices of servers, as far as the HCA goes. It was either all-new certified Dell EMC equipment or equipment from another company that they can place these servers on. With our being an all-Dell EMC shop, and my being familiar with Dell, I opted for all-Dell EMC hardware.
Being a small business, we don't have another product alongside it. It is the product. So it gets 100 percent usage. I don't see us expanding our usage in the future. The power and the storage should last us for, hopefully, the next seven years, which is roughly the Dell EMC support contract life expectancy. We use our servers for seven years and, at the end of the support, we refresh and buy brand-new servers.
Nothing stands out, in terms of problems or issues. They helped me and got everything resolved that I had problems with. I would give it a 10 out of 10.
Which deployment model are you using for this solution?
Which version of this solution are you currently using?