Dell EMC PowerMax NVMe Review

The compression and deduplication are the most valuable features because of the cost savings


What is our primary use case?

The primary use case is the database and high transactional use for block storage.

How has it helped my organization?

It simplifies things, as we are using higher availability systems. We have always had a 70:30 effective rule between vital critical applications and those systems which are not critical, important, and discretionary. The Symmetrix's line has always been the go-to for our vital, critical application types. 

We are also implementing storage virtualization with VPLEX, where it is giving us the ability to move the storage to the proper platform based on the operating characteristics needed by that platform.

Symmetrix has always been a high-end, high availability system where we have never had one go down. I have never known one to go down. It is just an operational-stable platform that we still have to manage and maintain. However, it's not like some of our lower-end systems, like on the CLARiiON side of it or the Unity systems, where we have to be careful what we put on them and what we do with them because of the loading of the systems. 

This is why we need to virtualize the storage element, so we can move things to the proper platform. We are going to find that we will have more systems going from the mid-range to the higher range, specifically because of the needs of the platform. Wherein before it was driving by someone determining cost for what they wanted to pay for the storage, which overloaded and created problems for the lower-end tiers, because they're pulling too many IOPS out of that tier.

What is most valuable?

The compression and deduplication are the most valuable features because of the cost savings.

What needs improvement?

There are glitches in the system at this point in time. I don't get the information that we've traditionally gotten, things like the power report that used to be standard for all EMC components across the board. Now, you can't get them. You get a little bit here in that report and a little bit in another report, but you never get the total amount in one report which gives you the equipment, its power utilization, maximum recharge, the interfaces for the power, and what are the requirements for the interfaces on the other end, so you can know exactly what has to be connected at that point.

For how long have I used the solution?

We just started implementing the PowerMax 8000. We have them in a couple different data centers, and we are looking to put in another three in other data centers.

What do I think about the scalability of the solution?

So far, the scalability has been good. Normally, what we end up doing is putting in a base platform, then adding capacity. However, the capacity adds appears to vary 30 to 45 days on us being able from a time of request to a time we are able to implement, because we were using Cloud Flex. We had been on a utility model for almost six years with Dell EMC, who was one of the early adopters of that type of technology, and in the first three to four years, it saved us over $5 million in expense.

It's the whole thing of having to go out and buy a multibillion dollar array. It takes us so long to ramp up and be able to get on it, due to the lifecycle components and because of the storage virtualization side of it. It is just incredible how we can just roll in a new platform, transfer the workloads, and have it up and running in days, instead of months.

How are customer service and technical support?

We live with technical support. We employed a platform called ViPR SRM, which I think they are redoing. It allows us to take a single pane of glass for an entire storage infrastructure. This gives us a window into what's going on and problem spots that we need to work on. However, there are always times when something is breaking and we have to work with tech support to get it up and running. There have been a few cases where tech support didn't work out well and other cases where they met the margin. 

Since the merging of EMC and Dell into Dell Technologies, there has been a hurdle that they've had to overcome, and they're not over it yet. It takes two to three times longer for things to get fixed than it did when they were separate companies. That is something that has to be fixed.

If you previously used a different solution, which one did you use and why did you switch?

Since 2016, we have been buying all-flash components and the price has been trending downwards. What we are seeing in the new products, since we went from the VMAX3 to Power Max, is the price point still drops on the overall cost of the storage. This is what we're trying to do. We're trying to get more value to our internal customers by reducing the cost of usage.

Our performance requirements were response time and what IOPS we needed out of the platform. It's exceeding what we're asking out of it when we looked at the PowerMax which we have already deployed. We are getting typical response times in half a millisecond (or 500 microseconds) lower because the target was supposed to have been the Generation 3 with 300 microseconds. We were supposed to be able to get around 250 microseconds with the PowerMax 3000. I have seen some of that happen on monitoring side. It doesn't happen all the time, but for certain applications, it does achieve going down to the 250 to 300 microsecond range.

How was the initial setup?

The migration from the VMAX arrays to PowerMax was done using our storage virtualization via VPLEX. It was just a matter of submitting the mobility jobs and keeping the queue full as long as it took to get that done. For the storage that was virtualized, that was done in a matter of days in transitioning from the old Generation 2 and Generation 3 to PowerMax.

What was our ROI?

On the Symmetrix's line, we are moving away from the VMAX2 and VMAX3 that we have and moving to the PowerMaxs 8000, as quickly as possible, because of the financial incentives. Also, the cost per gigabyte is reduced by at least a third, if not by a half overall, mostly because of the duplication.

Which other solutions did I evaluate?

Right now, we're doing a PoC with HPE 3PAR storage.

We currently have almost 100 percent Dell EMC storage technologies in play. We are looking at it as other vendors to see if they have comparable products, what we can use, and if there are very similar things to what we have.

I've even been to other classes for other vendors, like IBM. I have looked at IBM storage for various applications and come back to say that we are doing the best of class, so far, because what IBM recommended for the solutions does not exceed the current platforms that we current use for those storage solutions.

Disclosure: I am a real user, and this review is based on my own experience and opinions.
Add a Comment
Guest
Sign Up with Email