Dell EMC VxRail D Series - What are the limits
Transcript
(upbeat music) >> Hey, how's it going? It's Keith Townsend, Principal of the CTO Advisor. You're watching yet another Intel sponsored video. We have Bill Leslie, Senior Manager of technical marketing engineering at Dell technologies? >> Thank you, Keith, a little bit of both. Technical marketing engineering, so I kind of cover both facets. The fun technical side of it as well as the fancy side of it, the marketing side of it. So we can go wherever you want to take this one.
>> So, we telegraph this that this is going to be a tough conversation. It's sponsored by Intel so I can go after Dell as much as possible. We're going to talk about VxRail at the edge. You prepared some material for me, but we kind of threw that out the window. >> Well, let's keep it real, right? So I'm just going to toss those slides out, have the conversation with you because that's going to get us to the place where we need to be with this.
So, what we've been doing with VxRail is really making it easier to manage at the edge. Having more options for different node types that can be deployed in very harsh environments. I think you joined us this summer when we launched our D Series, our durable series, smaller footprint, shorter depth, mil-spec certified, for those really harsh demanding environments, whether it's high humid conditions, high wind, high dust. Really some great ways to get VxRail into more harsh, rugged types environments. You can think oil rigs, for example, as one of those types of scenarios.
We've got a lot of customers that just need that smaller footprint. They can't have a full standard rack or standard aisle. They might only have a little bit of space underneath the desk to be able to deploy their HCI environment and VxRail can now do that with our D Series. >> So when I hear smaller, I think underpowered but there's a key difference. Yet, I hear oil rigs, responsive data, smaller, underpowered. Do I get the oomph that I need to collect the data and analyze that data at the edge with VxRail D?
>> That's a great question. So, power, the way that you're defining here is really about the compute processing power. And the great thing about our new D Series is the fact that it's using the latest and greatest second generation Intel scalable processors, right? So there's no sacrificing in terms of the compute performance that you're going to get with these. Still fully capable of doing the all flash environments. It's really a great way to really just think about how do I shrink down a little bit of that footprint, I can use some of the larger capacity flash drives in it as well.
So I get the best of the compute powers as well as the density of the storage in these platforms. >> Well, talk to me a little bit more about this Intel flash capability. What have you guys done special with Intel to optimize flash at the edge? >> Sure, so one of the things that we're just bringing to market is the Intel Optane PMem, the persistent memory, and specifically in the memory mode, this is really great especially for those edge scenarios because the cost points or the price point of these is far, far better.
You're able to actually, for about the same price as traditional DRAM, get about 25% more workload density out at the edge. So it's a great way of utilizing these technologies to actually increase how much workload you can do. So you're getting that compute power. You're getting some extra density in the workloads, as well as the storage capacity with the flash that goes into them. >> So I'm going to push back a little bit on that. I was at VMworld 2020, virtual event, we were part of the Holstein or judging of Best of VMworld, until Optane won Best VMworld for innovation.
However, some of the innovation from some of other judges is that it's early, especially this PMem technology. We're hearing a lot about in-memory databases, et cetera, but how does that practically help today? Is it a little bit ahead of its time? >> So I think it's important to remember that there's two different cases for Optane PMem, right? You've got the application direct mode which is really the one where you need to have that application stack that can optimize it. The second being the memory mode.
Memory mode is, think of it more as an extension of your DRAM memory, right? So it's actually taking advantage of almost a memory tiering type of capability within the vSAN environment, where it's pulling from that much denser higher capacity Pmem that's at the lower cost point. And because of that, you're now able to get more workloads running at a similar price as what you would if you were to be doing all DRAM, for example. So it's a great win on the memory mode.
In the application or app direct mode, specifically, it's a little bit different. We also do support that. But in those environments, you're now dealing with persistent memory as a storage tier rather than a memory tier use case scenario. So now you've got to have that application that's actually protecting the data. That's doing the things like replication, for example, rather than relying on a software defined storage, like vSAN, to be doing that type of protection level. So in those scenarios, absolutely, you got to make sure that your application stack is ready to support it.
>> So VxRail, as a brand, I think about that as the heavy solution for my data center. It is the thing that I run everything in my data center from SAP HANA, et cetera. Give me some examples of edge specific workloads that we run on the series D. >> Sure. I think it really boils down to how easy is it going to be to manage. And, at the edge, you probably don't have the same skilled resources that you do in your core data center, right?
You might be dialing the friend, phone a friend back into the core headquarter location for help with what's going on. The nice thing about VxRail is all of our management is built right into V center. So it's the same look, same feel, same type of workflows that we have. But we're automating that full stack of the VMware environment as well as VxRail through our HCI system software. We also do some things with our SAS multi cluster management. This is a framework where you can now manage numerous VxRail clusters from a single location.
So, let's say, again using the oil company scenario where you might have hundreds, thousands of rigs, you can now view all of your VxRail clusters from that single source. You can even go so far as to stage updates to all of those clusters from one pane of glass. And we've just recently added the ability to actively update through that same interface with our SAS multi cluster management capabilities. So somebody that's back in the central data center operations can actually manage hundreds, if not thousands, of VxRail clusters, and make sure that they are all maintained the way that that particular company or IT staff desires it to be.
>> So I get the value of VxRail over, like, a Ready Node. I can't do that with Ready Node's lifecycle management is a great invaluable, but we're talking about the edge. The managing physical devices at the edge is tough. Talk to me about the practice goals of like expanding the storage systems or actually just expanding the system. What am I getting myself into? >> Well, the nice thing about what we're doing with the VxRail is we're automating a lot of those types of operations.
So when it comes to something as benign as needing to replace a drive, quite simply, unplug the old one, plug the new one in. Everything's auto discovered and incorporated back into the VxRail cluster. There's not a whole lot of expertise needed. If you need to add a new node, for example, you plug the new node in, it's auto discovered by the VxRail. It says, "Hey, I'm recognizing there's a new node in this environment. " And, by putting in the proper administrative credentials, of course, that node is automatically built into the cluster.
There's no special expertise that's required. It's quite simple in just a few steps, and really initializes in a matter of 15, 20 minutes type timeframe. And you can actually add up to six nodes in that single batch environment. So, very easy to scale as quick as you need to in that environment with VxRail. >> Bill, I appreciate the conversation. Did I miss anything? >> Yeah, Keith, there's one area we didn't talk about, and it's one of the hot topics that the VMware just announced at VMworld.
That's the ability to run cloud foundation where you've got remote workload domains in a different location than your core management. >> Look, hold on, I'ma stop you for a sec. VCF is the heavy part of VxRail, you're telling me I can run VCF all the way out to the edge? >> Absolutely, can be. That's the beauty of VxRail. The same nodes that are VxRail are the same nodes that you can run in cloud foundation scenario. So now you can actually have these remote workload domains, using the D series out at that edge, all being managed from a central management cluster and domains set up that's out of the core data center.
Very, very powerful to get all of that networking security with NSX, as well, as a part of cloud foundation. >> So, I get the Kubernetes distribution, Kubernetes grid, as part of VCF. So you're telling me I can run Kubernetes distribution all the way to the edge? >> That's right, as long as you're licensed for it Keith. You can run it out at the edge all day long. >> Wow, Bill, I love the breakdown of VxRail D Series out at the edge.
I learnt something today. I hope the audience learnt something. If you have some feedback or even pushback, VxRail is that too heavy of a solution for the edge? Can you do better? Hit me up on Twitter @CTOAvisor at Twitter. Thank you Intel for sponsoring this video. Talk to you next CTO adults. (upbeat music)
