The most valuable feature is that it's an end-to-end solution for which we have a single support structure. We can tag it in the event we have something that crosses between silos.
We like that we've got the validated design that we can use quickly to start it. We have something that we know that will work. We also like the fact that the whole end-to-end design has been tested. And we know that if we have a problem, we're in good company possibly. That's a big help for us.
Improvements to My Organization
The benefits to my organization are faster time for deployment, faster time to problem resolution; a stable and reliable environment.
Room for Improvement
I think it still seems that there is isn’t end-to-end automation. I can't say, "I want to build a host," and say, "Provision the blade. Provision the storage. Provision the SAN." All those things in one swipe would be nice. I think that's what everyone wants: push the button, make it work. The big thing is that we'd like to be able to have it provisioned end to end. The biggest weakness is on the SAN side and the fiber channel side is a mess. It isn't very clear at times. That part is the most complex. If we can avoid it, we don't do it.
It's been a very stable environment for us, absolutely. We've only had very few instances and each of those have been resolved very quickly. It's been a fantastic change for us.
It just works, scalability-wise. We've been able to continue to grow the environment. We've tripled in size since initial deployment. We've not really changed the overall structure of the deployment. Scaling up was very easy.
Customer Service and Technical Support
Technical support has been very good. We haven't had to use it much, which is good too. We've had once incident where we had to really get all three legs involved – the compute, the storage and the network side involved – and it worked well. It was a very good experience to see from end to end. The problem looked like it was storage, but could've been on the SAN side. We worked through that and we also had to go back through the host. We had all three parties involved. They worked well together.
Initial setup was very easy. We looked at the validated design. We tweaked it a little bit to fit our environment specifically. We deployed from there. It was a day or so at best, more because we have our own little processes we have to go through.
ROI is difficult to gauge because things have changed so much in the time since we started. We have moved from a virtual environment of roughly 30, 40, 50 percent to 70, 90 percent now. I want to cut it.
Other Solutions Considered
We've been using HP for a long time.
For storage, we were using EMC before. We had been using their platform for a number of years. We were in for a forklift upgrade. The next upgrade would've been a complete forklift upgrade. We evaluated Hitachi, NetApp, and EMC. We talked with Dell as well, but we really narrowed it down to Hitachi, EMC, or NetApp. With EMC, because it was a forklift upgrade, we were really upset that it wasn't something that we could just add on to what we had or change out the parts. We had a major conversion, so that put them at a disadvantage right away. In talking with NetApp, the big thing for us was getting rid of the forklift upgrade; going to a system where we could evolve from year to year to year. We're three or four years in on this now; so far that's been the case. We have done controller replacements and upgrades and everything else, without serious service interruptions, so that's been a big deal for us.
You need a partner that you can trust and a partner you can actually move forward with. You need to understand what the full lifecycle of the product will be. You need to understand where you're going to go next. That was the big thing for us. When we changed, it was a major change. It was a lot of downtime, a lot of frustration for a lot of people. That's why something like NetApp, where I know that I have the ability to take out the controller, grab a new controller with cluster data ONTAP; you've got the scale out ability as well. That's been really great for us as well. As I’ve mentioned, we've tripled the amount of storage, we've tripled the amount of compute, added additional controllers, with little interruption to users. That's huge.
Disclosure: I am a real user, and this review is based on my own experience and opinions.
Nov 01 2016