What is most valuable?
In a word, performance. It was primarily brought in to replace our EVA with something that was more capable from a performance perspective. We did not order solid state in our original configuration and we were looking for something which could grow with us, that could handle unpredictable VMware workloads better, and that didn't have the bottlenecks of a traditional monolithic array. Since that time, we have added solid state to accelerate the performance further.
What needs improvement?
My 3PAR array is too old to support File Persona, and I would love to be able to do that, but that's something that we would need more powerful controllers. We've got the first generation, so at some point we will get it, but we've got to wait for a refresh cycle.
What do I think about the stability of the solution?
3PAR stability has been there since day one. It's one of the platforms that has just been a joy to work with, because it changed the way that we are able to protect our entire environment. Being a telecom, we've got really good high speed links between our two data centers and we are able to do with 3PAR in a Peer Persistence configuration, and that's a feature where it clusters the two 3PAR rays together with seamless failover for LUNs from one array to another.
Now I've got disaster avoidance, rather than disaster recovery of a traditional replication technology. So Peer Persistence for us is kind of nirvana. It's been a great solution for us.
It started out just with VMware, then they added support for our Windows clusters. Most of the things we run on the blades are things that we can do this with. So if we have a blade fail, the great thing about a blade is the server profiles, move that to another blade, spare, in the enclosure, it comes back up in the same server running again, and if we have a storage failure, it automatically switches over in the backend, and our users never know.
What do I think about the scalability of the solution?
We've got plenty of room to grow. We're about three or four years into our 3PARs and we've still got more than at least about 50% of our drive shelves open, so we have a lot more room to grow. With each generation of drive that comes out, we can install bigger and more capable drives in it, so we haven't hit any scale issues there.
How are customer service and technical support?
7/10 - it can be hit or miss. We get better luck with our premium support levels. We have a named TAM for some of our systems, that works out well. Escalation managers are always good. There is good technical talent, it's just sometimes hidden by first level support. That can be difficult and frustrating at times, but over ten years working with them, I would say today it's probably a little better than when I first started. Actually, I would say it's probably improved a good bit since I first started working with them, but it's still got some room to go.
Which solution did I use previously and why did I switch?
We were using HP EVA, and before that we were HP HSG. So when I first started we had two small HSG arrays that were primarily behind DMS and maybe a couple of other systems, but we had a very small amount of data on a san. VMware changed all of that. So we had all of this data running on EVA and we were staring to hit some limits, and the EVA didn't have good telemetry to let us know where the edge of the cliff actually was, so we were teetering right on the edge and about to fall off when we got the 3PARs in. So that kind of saved our bacon.
When we were evaluating potential replacements for our HP EVA storage array, we looked at Tintri and Tegile. We ended up doing evaluations for both of them.
Our company purchased Tegile for another project. Internally, my security officer didn't like the NFS of Tegile. That was kind of a no-go for internal use. Otherwise it has really great features for virtualization. That was really appealing to me as a VMWare administrator. We talked with them after a VMUG meeting and brought it in, did a proof of concept on it. It didn't perform as expected in our environment and we found out after the fact why. We would have needed a second active controller and a second disk shelf to get the full IOPS we were expecting out of it. That kind of killed our evaluation. It had good features, it had good reporting, which was one of our big criteria moving from the EVA. We wanted something that was going to let us know how it was performing. That was really strong, in Tintri. It's also been three and a half years since then, so that product has changed a lot also.
How was the initial setup?
3PAR installation was an easy one. We had really good consulting services that came in and walked us through that process. We've done a couple of field upgrades and those have gone smoothly also, so 9/10.
Which other solutions did I evaluate?
From a storage perspective we've looked at some other vendors, but once the 3PAR 7000 series was announced with its capabilities, it made the most sense, being mostly an HPE shop.
What other advice do I have?
I really love that platform. It's rocked for us. It's, like I said before, it's near nirvana for our environment, because we are trying to do something where we want to avoid disasters and have seamless fail-over, and I don't know of another solution that can accomplish that in our environment.