What is our primary use case?
Right now, we are finishing the proof of concept stage. We're using it as a data store for a virtual infrastructure for telephoning applications.
Thus far, it's met all of our expectations and has actually has exceeded them a bit. Once we got it up and running, it seems like it runs pretty rock-solid. There have been a couple hiccups here or there related to how some of the fault tolerance features work. But overall, it's performed very well.
How has it helped my organization?
The hope is right now it's going to empower us to lower mean time to resolutions. Since it's being used to support a virtual infrastructure, that gives us access to all of the benefits of a virtual infrastructure. So, instead of spending hours trying to rebuild or to fix a broken server, we can instead wipe it out and throw up a new one and we're up and running again.
What is most valuable?
I like the speed capabilities that it provides. The deduplication features definitely have some huge potential. The latest firmware, where they've enabled compression for workloads that aren't very good for deduplication, I can definitely see huge potential there. It's just going to be a matter of how much time we have in the lab to test all of them.
What needs improvement?
Automated reporting, real-time reporting would be excellent. Right now, most of the reports are ad hoc, so I have to say, "I want this report." I believe you can schedule them and have them go out to an email address, but it would be nice to have an actual dashboard where I can see in real time, "Ok, this is what everything looks like." It may be there and we haven't seen it yet. It hasn't been the primary focus. We're still in the process of engineering a solution.
For how long have I used the solution?
What do I think about the stability of the solution?
The "hiccups" I mentioned above did not cause downtime in any environment of consequence. It went down when we were trying to break things. When we tried to break it, it broke. But thus far, otherwise, it's been pretty darn solid.
What do I think about the scalability of the solution?
I can't really comment too much on the scalability. Most of what we're using right now is still a relatively small deployment. I don't see anything that would cause a problem with scaling outwards, and the system is made to scale. We're probably going to be doing our first augment next year, but we haven't really focused on it yet.
How is customer service and technical support?
We've been working directly with our product management group to work through any issues that we've encountered. There has been some stuff, but it's mostly related to it being a proof of concept, or the gear being proof of concept.
For the most part, whomever we've been dealing with, they've been knowledgeable and helpful.
There were one or two small issues where it would have been more helpful to talk directly to a subject matter expert. We're an engineering organization, so sometimes it's very difficult and we're asking for a very specific thing, and sometimes not talking to a subject matter expert right away can be frustrating, it feels like you're getting deflected a little bit.
But as of today, most of those issues have gone away, and we've gotten answers to all the questions we've had, eventually, so overall it's been a very good experience.
Which solutions did we use previously?
We worked hand in hand with our product management team to try and determine what the best fit was. In the past, we worked with HPE MSAs and HPE P2000s.
This was a logical step forward without going into any really crazy high-ends like Hitachi or EMC type of SAN which has all of its own problems that you get when you go in that direction.
How was the initial setup?
The initial setup, when we first got the gear, was actually done in coordination with HPE. They had installers on site. We didn't have that much to do. The next parts of the setup went very smoothly, integrating with Synergy.
Which other solutions did I evaluate?
We didn't so much evaluate other vendors. For different types of workloads we're looking at some open source solutions like Gluster, Ceph. There are a couple of other options we've looked at.
What other advice do I have?
Our most important criterion when selecting a vendor is support. We're in the telephone sector, or carrier sector. If we have an outage it's going to impact millions of people. So sitting on hold for 15 minutes isn't acceptable, especially when it could mean 911 calls aren't being made. That's really key, we need to be able to get to someone who knows what they're talking about immediately. So far, we haven't had that huge of an issue with HPE hardware.
I give it an eight out of 10. Still extremely good. Way better than most. A couple rough edges that could be buffed out. We're about 90% of the way to where we want the solution to be. When we're there 100%, and we have it working exactly how we want it to, I think my score would be a little bit higher.
I would tell a colleague who is looking at similar solutions that this is a worthwhile investment to look at, so long as you have the ability to test it appropriately and make sure that it works for your workload. Read through all the best practices, make sure that it will work for you before you try and do it, so you can actually give it a fair shot. But overall, give it a shot. I would recommend a proof of concept.
Disclosure: I am a real user, and this review is based on my own experience and opinions.
Dec 26 2017