Podcast Archive - StorageReview.com - Podcast #149: Dell PowerStore Elite: Hands-on Deep Dive From the Hopkinton Lab

Episode Date: May 29, 2026

The Storage Review team visited the Dell campus in Hopkinton to go hands-on with the latest advancements in enterprise storage. This podcast takes place in one of our favorite places, the Hopkinton Dell Lab. Brian connects with Scott Delandy, an Engineering Technologist at Dell, for this interaction to go in-depth on the PowerStore Elite. This The post Podcast #149: Dell PowerStore Elite: Hands-on Deep Dive From the Hopkinton Lab appeared first on StorageReview.com.

Hey, everyone, welcome to the Storage of View podcast. Scott, thanks for having us to the campus today. Great to be here. Very excited for this one. This has been a good day. We've been in the lab all day, which means we're actually doing stuff and touching gear and not just looking at slides. That is so old school to be hanging out in the lab, racket and stack in, looking at blinking lights. Are you even badged into the lab?

Can you get in? Oh, geez. I can get into some of the labs. I couldn't get into the lab that you guys are into. because that's the super secret lab. Right. So you have to be, you know, part of the bubble in order to be able to get in there. Wait, so they'll let you sell power store, Gen 3, but they won't let you actually go touch them?

The lab that I can get into, there's plenty of those things laying around. So yes, so I can see them anytime I want. So sometimes, you know, if we're in the neighborhood and I've got the family with me and we're looking to kill some time, we'll come and we'll sit in the lab and I'll say, guys, that's a power store. That's a PowerMax. That's a PowerFlex. There's nothing I could imagine your family would like to do less. then come to the lab and look at hardware. Well, you're probably right, but they humor me, so they're good about that.

Well, that's handy. I can't complain. So we've been talking all day today about PowerStore Gen 3. It just dropped at Deltech World. It's no longer Power Store Gen 3, so we can now use the official name. It is Power Store Elite. Elite, okay. Yes, yes.

So last year at DelTech World, we did the Power Store Prime release, just to give you the math on that. That included the new software of the 4.0 Power Store OS release. This release of Power Store Elite includes two things. It includes the new Gen 3 hardware, the platforms that you guys got to play with today. And it also includes the new Power Store OS 5.0 release. So there's a number of new features and capabilities that are included in the software that aren't necessarily tied to the new platform, the new Gen 3 hardware.

There are things that support the existing platforms as well. So there's a lot of backward compatibility and investment protection included in this release outside of the new hardware platform. The hardware is unmistakably exciting, though. We could talk OS5 too, and that's fine, but the hardware is huge. Hey, I've been doing this for a long time, and I love to geek out when it comes to storage hardware and technology and all the different things that are going in. But yeah, you're absolutely right.

It's a really significant refresh of the technology and kind of the guts that make up the storage platform. So all the latest and greatest things from a component perspective. You've got your PCI Gen 5, you've got your DDR5 memory, you've got the new E3 flash media, you've got the move to OCP. I-O modules. So, yeah, a bunch of really cool things that are included with this that, you know, quite frankly,

take a lot of time from an engineering and a development perspective. So these are major releases that, you know, I can't even tell you the, it must be, you know, hundreds of thousands, if not millions of hours of man time, you know, human time spent in not just the development, but the testing and the QA and just getting everything ready to be released and introduced for customers to be able to take advantage of. Well, you mentioned a lot of the core components, the E3 Flash being the biggest when you support the S and the L modules. So the design of this thing, if you looked at it from the front, I don't want to hurt your

feelings, but it looks like a power edge. Can we concede that point? You're absolutely not hurting my feelings because I think if you were to take the covers off and take a look at some of the components. We are using a lot of the standard technology that has been available in Power Edge for a number of years now, right? It's the newest versions of those components.

But that is done by design, and it's a part of the overall strategy because, Brian, I don't know if you know this or not, but there's a bit of an inventory issue around components and supply chain. Is that right? So having the luxury of, of being able to leverage common components

across the ISG portfolio gives DELA a very big advantage when it comes to supply chain management, meaning that when a user orders a system, we're looking at the ability to ship within weeks and not months or quarters in some cases. So I think the idea of being able to simplify the platform and being able to have a common parts bucket, if you would, that we could leverage across different platforms

within the portfolio is a huge strength that you get at Dell. And it goes to the scale that we develop technologies across, right? I think that's a big advantage that maybe a couple of years ago wouldn't be as intuitive, but I think just because of the way the market is right now and just all the constraint around all these different components, it definitely becomes kind of a key benefit in terms of what Dell can do.

Well, it's through the whole chassis, too, what you're talking about by standardizing on components. We've got Kevin nerded out with your lab guys. Yeah. It tore the controllers apart and looked all the pieces. There's nothing in there that's hard to find. Like, the... So if we look at Power Store Gen 2, for instance, like the I.O. modules are bespoke for that platform. Now you're going to OCP modules that are...

I mean, you sell probably millions of those across the rest of the whole portfolio servers and storage. So those are easy. And the drives, you can do E3S, E3L as customers ask for larger drives. You just have to call the part. It's not like you have to go design something or find some unicorn to shove in there. Well, there's a little bit more of that, right? I don't want people to walk away from this thinking that we're just taking off the shelf components

and sort of cabling them together. It's a simpler design, I would argue. It's a simpler design, but within that design, you know, it is still an engineered storage platform. So this is not an off-the-shelf power edge server that we're rebranding as PowerStore. It's still a two controller platform in the back, the HA availability, all those things that have been integral to Power Store over the years. Correct.

Yep. But there's a level of optimization because when you're designing and building a compute platform to do storage, it's very different than designing a design. compute platform that is doing kind of generic server type operations. So you're able to optimize the way the processors work and how it's accessing memory and how things are connected over the PCI, internal buses, and all of those things. So there's definitely a hardware element that has a lot of innovation built into it by taking these off-the-shelf components and being able to leverage those.

The other piece of it, and we mentioned a little bit, is the software piece. because again within the architecture, there are a lot of things that we're building into the platform that users are going to see immediate benefit from. You're going to see much higher levels of performance. You're going to see a lot more autonomous operations with things that are running inside the platform itself, as far as AI ops, as well as things that sit outside the platform

to help around the management for the entire ecosystem stack. But it is that combination of the software that sits on top of the hardware that is really the innovation piece. And really what differentiates this from the Power Store from other things that are available in the marketplace today. You mentioned a couple things there. So OS5 comes out with this system in ships in the summer sometimes. July. Okay, in July. Yep.

So that'll be there. I know security is a big deal to customers because we're seeing it. a lot in the data protection space. So when we look at Power Protect, for instance, they've got a whole different take on cybersecurity resiliency, being able to come back from a challenge there. But primary storage has a similar set of security objectives. Sure.

And I know that gets hardened in OS5. Yeah, so there's things that you're able to take advantage of on day one, right? So things around, you know, the big concern, that I hear from customers when the discussion revolves around cybersecurity is, what are we doing around ransomware? How are we securing the environment to minimize those types of threats? But more importantly, what are we doing to build in the detection capability? So if there is an environment that is compromised and ransomware has been injected,

how quickly can I discover that? And then when I do discover that, what's my path to getting things back online? What's my remediation for that? So how do I protect, not only do how do I protect and detect against ransomware, but then how do I go ahead and recover? Right. So within the platform itself, we're doing some really cool things around anomaly detection, being able to look for particular patterns or signatures that are indicative of a compromise as a result of ransomware. But then not just being able to detect that faster and more accurately because you don't want a whole bunch of false positives going off all the time because then it's like the car alarm, right? You ignore it.

The car alarm goes off or you just ignore it. So there's that sensitivity that you're always trying to tune. But then what happens when there is an event and now you need to recover? So how do you reduce the recovery point objectives, right? How quickly did you discover that there was an issue and then how quickly can you recover up against? that. Are we talking? Did it take hours or days or even weeks to discover that, you know, a piece of ransomware has been writing a bunch of encrypted data and, you know, we just, the user wasn't able to catch it because it wasn't something that they were actively scanning for.

And then once that becomes apparent, then how quickly can you recover and where do you recover from? So a lot of the things that we're building into Power Store, and not necessarily specific to the Gen 5 hardware, but just across all of Power Store, these are things that we've been actively developing because it is a key sort of concern that customers have. What the architecture within Power Store gives us the ability to do is to provide a lot more granularity and a lot more sort of telemetry that we can now monitor not just an entire application, but we can look at the individual I.O. that's running to that application. So in order to do that, you need an architecture that can kind of handle that because while you're

going in and you're running this analysis to see if there's anything funky going on in the environment, you still have to service I.O. You still have to maintain the data services that are included in the platform. You still have to do all the things that the storage array does, but you have to layer this additional capability on top of that. So architecturally, the platform provides a lot of things going forward that we can begin to leverage that takes that recovery point. When did you discover that there was a ransomware attack happening from potentially days to weeks and get it down to hours to even minutes? So that's the direction that we want to be able to go. And again, it's a part of the design of that architecture that is going to enable

us to do some really cool things on day one, but continue to advance and innovate in that space to be able to close those windows around RTOs and RPO's. Well, you talked about that and you mentioned AIOps, and we should talk about that a little bit more because some of these things really blend together, right? Because you're doing this detection, this monitoring, this intelligence to prevent data attacks, but at the same time, you're taking additional tools like AIOps and making it easy, to bubble that information up to the practitioners? Because I think we've got this challenge that there used to be a storage admin team.

Yep. And in many works that still exists and will exist forever. But in most, I think you could probably argue, the roles are getting stretched. You have more generalists involved in the infrastructure. And anything you can do to automate or at least automate suggestions, I think is absolutely huge. I'll give you, so, the perspective that I've seen. So I've been doing this for a long time, right? So we were talking a little bit earlier. So, you know, I came in through Dell from the EMC side of things, and I started, you know, back in 1990. So I was 11 when I started. So if you're

trying to do the math, right, I'm still pretty young. So I've seen a lot of things in storage. And, you know, years ago, when you started to talk to users and admins about automating a lot of things, especially when it comes to, you know, making changes or making updates, or moving things around, you know, the storage people are pretty conservative, right? Conservative is putting it lightly. Don't touch my stuff, right? It's kind of the thinking of that. And so, you know, initially the conversation was, yeah, you know, it's okay if it can help me

and it can make suggestions and recommendations. And over time, you know, the user begins to kind of trust the infrastructure to be smart enough to get it right. And I think what we're seeing now is. is kind of that shift where people are like, well, you know, I still want the knobs, I still want to okay, I still want to approve, I still want to have sort of the final say

before the infrastructure automatically does something. I'm seeing a shift where users are saying, you know, okay, I don't need to approve every little thing. You know, if it's a software updates, if it's a security patch, if it's something that is not a major change that could have significant impact into the environment, I want the platform to go ahead

and do that automatically, and just let me know that it happened. Right? And I think years ago, you know, you didn't quite have that level of trust from an admin perspective to let the machine kind of do its own thing to whereas now just because, you know, a lot of things, a lot of these applications that are running in the cloud, there are no knobs, there are no levers, there are no alerts. There's no screen telling you to do thing. It just automatically happens for you.

And I think that's really kind of reshaped the user, experience and really the expectation around how they want that infrastructure to manage going forward. So when I talk about AI ops with a lot of the users that I work with, you know, whereas before it was like, yep, just put the safeguards in there so that I can make sure that, you know, I can check the box and things will move forward to today where it's like, no, you've got to go do it faster, do more, here are the use cases that I have that take up a lot of time and a lot of admin effort. And if you can automate those things as much as possible, that frees me up to be able to do things that are more productive

and more value to the organization. But it's been a shift because I think years ago, there was definitely resistance to that because people just did not trust the machine. And I think just because we've been iteratively increasing the functionality and increasing the automation over several years and in many, many releases, people are starting to become more and more comfortable

and now, in fact, do trust the machine to in many cases make better decisions than what an admin would be able to do just because we're not looking at that single system, we're looking at the entire install environment. And we can kind of predict and we can be more prescriptive in giving recommendations to write to the point where we're automatically doing remediation for you, where, again, the user doesn't have to do anything in order to make that change. Well, on the AI ops front, it's going to continue to get better, too, and more integrated and more information and give people more actionable things. And I'm sure there'll be some

settings to enable like at this tier of decision making, go ahead. At this tier, let me know. And at this tier, let's have a meeting about it and discuss. That's exactly it. And what I would say is that the things on kind of the left side that you just described are becoming fewer and fewer and the things that are happening over here are becoming more trusted where, yeah, there's fewer meetings, right? Do you think some of that is the catering? change in software releases? Because if you think back to your early EMC days, software releases for primary storage were like scheduled events and we really had to plan for it. And we did it Saturday at 3 a.m. or something to really avoid any potential downtime or risk to operations.

And they didn't happen that often. Now the release cadence on Power Store, for instance, the point releases, it's what, quarterly or maybe every... It's about three a year is what we're up to. Okay, so three or four a year. Yeah. And I think customers are, and it's been that way, and customers are kind of used to, okay, you're going to do this and I'm going to get XYZ new features. The features are substantial.

It's not just like sparkles. It's real stuff. The way the releases work is that you'll have a new release that will come out and they'll be the typical kind of patches and security fixes and those types of things. But there are going to be a number of new features and capabilities that are also introduced on top of that, which is what you want in a release. You don't want to just be going out and fixing bugs all the time. You want to be able to add new capabilities. And believe me, we've got a whole roadmap of things that are kind of in the pipeline for where we want to go for the next several releases as we begin to plan those out.

But what we don't want to do is kind of force the user every three or four months to have to upgrade code. So we've gone to this long-term support cadence where every couple of years, there's what we back in the day we'd refer to as target code. Here's the minimum recommended level of software that you should be at. And if you're at that code, any additional patches, security fixes, those types of things are going to be made available to that version of code. So if you're there, you're happy, you're stable, and there's nothing in that next release or that newest release that's come out that is interesting to your environment, there's no need for you to be forced to go ahead and do that upgrade. So it helps with users really kind of plan out their life cycle management around, you know, how often and how aggressively do they want to do these code updates.

And, you know, we've got some users out there that want to be on the edge and, you know, when a new feature comes, comes out, they want to bring it in, they want to test it, and they want to get it into production. And we have other users out there that are less aggressive in wanting to do that. And, you know, when things are stable, it's let's not go in and make any major changes until we're ready to go ahead and make those changes. You talked about EMC. I was thinking back to, and I'm sure you'll remember this, I think one of my favorite products, and you might think this is funny, that EMC ever... I probably have the slide deck. That EMC ever released is not some big V-block or VC or any of that stuff.

VNXE, I think, was one of the most important products that company released. Not only one of the most important. Oh, you agree. I do. I think it was one of the more misunderstood platforms as well. How's that? Well, because the VNX platform for file was based on a 32-bit file system. And the VNXE had a 64-bit file system, so it had kind of this next-gen.

But it was kind of positioned as an entry-level version of the VNX, even though the file capability. It was the E because we had other things in the portfolio that when we had a smaller version, we didn't know what to call it, so we just put an E on it, and that stood for entry-level. It's like a script-ee or something, too. It was kind of... But it was misunderstood because that file... capability became really the foundation of what we run file on today for the primary storage platforms.

So if you look at like a PowerMax or even Power Store today, it's not the same code, obviously. Sure. That's an old version. But that did become kind of the architectural foundation that the newest levels of file code were built on. So again, misunderstood because people thought, oh, it's this entry level thing. but there were a lot of powerful capabilities as well as, again, a very flexible architecture that allowed us to kind of go ahead and build on top of that.

So I remember it well and I do have the slide deck so I can email it to you. I don't need that again. But I remember we did, we got the array. It was it 3,200? I can't remember the, okay. Was that right? Yeah, that was it. Yeah, and that was a lot easier to use too than some of the prior product and really

It's funny, because the stuff we're talking about today with Power Store is a much more refined version of the conversation that was happening a decade ago or whatever it was for VNXE. Yes. In many ways. Yeah. I think when you look at, you know, this platform in particular, I mean, we are primarily, you know, installing these into the enterprise environments.

So when you look at the data services and the resiliency and, you know, all the capabilities around, you name it, scale, performance, security, you know, these are very mission-critical workloads that users are using this platform on. And I know it's one of the misnomer's out there where even, you know, just last week I was talking to a customer and they referred to kind of the product or kind of that space within the product as mid-range. And it's really not because mid-range, it's an incredibly inaccurate way of describing what this platform is because mid-range goes back to kind of the segmentation of the market

around price-banks. Yeah, we get locked in on price-pans. So if you're below a certain dollar point, you're concerned mid-range, and if you're above a dollar point, you're high-end. But it doesn't take into account the applications- You're being penalized, though, in some ways for being more affordable? Exactly. So it's, you know, it really is an enterprise.

dual controller platform that again when you look at the data services and in the maturity of those data services for some of those advanced you know replication topologies and being able to do active active failover and integrating that into environments not just kind of the ones that have been around for for a while you know you think of things like vmware and working with site recovery manager to fail over and fail back where people are going now is they're looking at you know not even running VMs anymore, they want to go directly to containers and use Kubernetes. So now in order to do a lot of the VME type of integration, you know, we're now using CSI or CSM, right, in order to do that, the container storage modules available from Dell that provide that higher level awareness of the

data services available on those storage platforms that allow you to be able to do the failover, failback, and do everything without ever having to go down to the power store level to make those changes or to do those that type of administration doing everything at the VM or really at an orchestrated level you know and again to your point around the AI ops having that that fully automated and not you know having any administrator involvement to fail over and fail back depending on the type of scenario that you know a user is facing that requires that type of an action well I was thinking back to again to the VNXE era and the things that were important

then are some of what you are I think I have a VNXC T-shirt. I want to go see if I can find one. You think it's in your truck out there? You know what? I have lots of T-shirts here, so I can probably go find one. So the challenges that enterprises were facing then, they've changed some, but you mentioned a lot of the blocking and tackling then, like replication and failover that still is foundational

today. But how is today's conversation change from what it was back in those days? I think people, I mean, back then, when people looked at a storage system, they evaluated that storage system based on that building block. Okay. So they compared this storage with that storage and what's the performance and how many gigabytes and what's the cost and all that? Where I think today they're not looking at an isolation of a storage array. They're looking at it as a part of the infrastructure that has to fit into the entire stack. So how does this connect into the host environment?

what types of applications are you running in there? What hooks are available on the application side to take advantage of some of the capabilities that you now have in these intelligent storage platforms? What are the automation tools that you want to use, right? CSI, CSM, if you're looking at containers or Kubernetes, for sure. But if you're doing something else and you're using Anciple or you're using Terraform, how does this play into that?

What are the playbooks and how easy is it for me to go ahead and just plug these things in, without having to go in and rewrite all of these scripts in order to make it happen. So I think that's more of the shifts. So a very common conversation when I, you know, talking to a customer about, you know, the storage platforms, again, you know, they want to understand, you know, what's the storage array and how fast is it going to go and how many drives and all of that. So those things are still important.

But the more interesting parts of the conversation is how does this plug specifically into their environment and how does this meet some of their requirements? around how they're set up without them having to go in and rip and replace or or build new things on top of what they already have in order to take advantage of the infrastructure. So again, I think that's the big shift. With VMware, it was pretty easy because everybody ran VMware. And as long as you could plug into the VMware stack, you were good. Well, there's some dynamics that have changed. So there's been some dynamics that have changed. So now the conversation is, it's not just VMware that's out there.

It's, I've got Kubernetes, I'm looking at Nutanics, I, you know, I have Azure in the cloud, and I'm thinking of repatriating some of those applications and running those as local. So how do I take your thing, right, and not just make it relevant to my VM users and admins and applications, but all these other things that I want to go ahead and deploy. Talk about that because that's important, because Power Store is so flexible, you end up having to be good at everything. And the world of everything continues to change and evolve. Obviously, you can't go far.

We've made it pretty far 20-something minutes. They did not say AI yet, but that's part of the conversation. You're there for all of those things. How do you maintain the flexibility but still execute well? I'm going to give you my product management answer. So I'm part of the ISG product management team. been in this organization for God, I don't know how long. It's been quite a while.

I think when we look at requirements and sort of what are the things that we need to do next, we work very closely with our users and customers and understanding how their environments are evolving, what things are shifting, what are the deployments that they're interested in moving to in the future, and then making sure that that's getting communicated back into the product. management in the engineering teams so that as we're planning out our roadmaps, we're doing it in a way that aligns to what our users are asking for. Because you can't just go out, and to your point, there's all these different options and all these different things, and you can't be, you can't support them all.

At least you can't support everything at once. At once and be kind of best in class in order to do that. You're spreading yourself too thin. So the way we do that here at Dell is we have a very focused effort of understanding What are the real things that a user needs? What are their pain points, right? What are the things that we can help them solve by doing things on our storage array?

And how do we go back in and how do we prioritize those? So you mentioned, you know, we're doing, you know, three or four releases a year, right? And we didn't do that a couple of years ago. We had one release every 12, 18 months. So now that we've been able to accelerate the cadence of the releases that we're putting out there, and adding new functionality into those releases,

that gives us the ability to very quickly kind of prioritize and reprioritize, right? Because when you're planning something a year or 18 months out, boy, you better hope that you're right. You better have that crystal ball that is a higher risk, I suppose. It very much is. But now that we're doing these releases a little bit differently, it gives us a lot more flexibility

to pull key things in or to maybe something that we heard from a set of users is starting to shift because there's other things that are happening within that space. So maybe we deprioritize some of these other things. But it gives us the ability to be able to kind of move things around between these different release vehicles and to do them and not have to wait 12 to 18 months in order to do that. We're talking literally weeks to months in many cases in order to be able to do that. So I would say to your question, how do you do all these things well?

The answer is you can't, which means that you have to be very good at picking the right things that you want to focus in on. And the way you do that is by making sure that you're connected in with the users and you're talking to partners and you're hearing, you know, what's the buzz on the street from, you know, analysts and folks that follow this

and making sure that you're getting, you know, the right feedback so that it can help with the planning and making sure that you're doing things and heading in the right direction. Well, I think the two directions that you're going right now are that are critical

are ease of use. So in intuitive GUI, the AI ops enhancements to make the whole system from day one on, really easy to manage and administer. And the second part is ease of service and deployment. And again, we talked about the componentry being able to easily service these things. I mean, even the boot drive got moved around

to make it easier to get to the event that service is required So that the downtime is minimal. And even this is one of the best things the guys are so excited about in the lab. Unless you've been in there and like really wrenched on these things, it's such a small thing. But the Gen 2 handles to remove a node kind of swing out and then you pull. Yeah. It's like opening an old window. Yeah.

But you could make a mistake if you weren't careful and maybe yank the wrong. Now it's got two buttons on these latches. It's very obvious what your service is. yank the thing out, do what you need to do, throw it back in. And you learn that by selling hundreds of thousands of Power Edge platforms and realizing, you know, one of the things that people do when they break, you know, something by either removing or replacing or upgrading. So how do you engineer that, oh, so that you minimize those types of issues?

And again, going back to the earlier point, you know, it's a good thing to be able to have access to that parts bucket where a lot of people have kind of learned, oh, we don't want, you know, that latch because that, believe me, I've been on the Deltech floor where somebody, you know, wanted to pop the cover off and we were able to pop the latch off and we've broken the latch. And so, you know, this is an absolute true story. And so, you know, we broke the latch and now we've got this thing and we can't put it back on. And it's like, boy, that was kind of flimsy, right?

And this is going back a couple of years. And then now we've completely eliminated it because you no longer have that flimsy latch. because you're not doing that every day. And you're hoping that the person that's doing it knows exactly what they're doing and the right amount of force to play to that. And you may never do it. But when you do...

You have us knuckleheads running around at the Dell Tech floor. Oh, you want to see the cover? I saw someone open up. I think I can figure it out. And you rip the latch off. That's not good. Yeah, the Expo Showfloor

is where all new hardware goes to get ESD to death or gets manhandled to the point of None of that stuff is probably going to operate again, but that's okay. And it all comes back and they'll ask the same question, well, who broke it? And we're like, I don't know. It's like my kids. Like, who left the light on? No one left the light on.

It just did. It just happened. It just did. Yeah, well, the platform, the hardware is neat, really neat. 40 drives in the front, flexible caddies so that you could do E3L sometime if customers want, these real high capacity systems, all the swapability components in the back. I mean, the Knicks of the front, anyway, we've got the whole deep dive on that. It's a significant refresh from a platform perspective because what this will do is it will set us up for the next several generations of technology to come.

Oh, speak on that drifts real quick. We talked a little bit earlier today with the guys in the lab about migration. Yeah. So for customers that are on prior gen power store, you could, I'm guessing this will cluster in and they'll probably migrate workload. but what's the migration path look like? So the migration path is very simple because there's a couple of ways to do it. The easiest way is to basically take the new appliance,

create a cluster with the old appliance, and then use the migration tools, the inter-cluster migration tools to be able to move the applications from the old to the new. And the reason why I say that that's the best way to do it is because the path management is all handled at the cluster level. So in some of the migration options that we've had, the past, once you've copied the data from old to new, you've got to go to the host and

you've got to change the past. Right, so you have to do a scan of the initiators for the host to be able to see the storage, but to see it through the new paths. By doing the internal migration through the appliance, you don't have to do that anymore. So it basically just walks you through and it allows you to move application by application or move the whole array. So there's a lot of throttling. that you can put in place depending on how fast you want to move from old to new.

And, you know, what I see is that in a lot of cases, once people move old to new, they still have that old storage. And depending on what the plan is for that, they're probably still going to use it for something. Sure. Right. So as long as there's some life left in it and, you know, they've got some warranty because, you know, we've made changes in terms of kind of the maintenance for those platforms going forward. So it's not as economically unattractive to keep.

those old platforms running but there's you know challenges with the older hardware just around kind of scale and performance and hitting some of the capacity limits so there's a need to go to the new hardware but there's again if it's if you need storage for some tier two tier three types of requirements you know have that I just had a customer absolute true story he said he still has a CX4 running in his environment and I had to look it up and I did have the PowerPoint presentation so it was announced and 2008 is when it came out and it went end of service life in 2016.

So he's been running it for 10 years. A decade past end of service. Now, I'm not recommending we do that, right? But I'm just saying the point being is that when that new thing comes out, it doesn't mean you take that old thing and throw it away, especially today with the challenges around supply chain and inventory and just everybody wanting to maximize the utilization of the infrastructure. that they have, you know, having that ability to move things from old to new provides a lot of

flexibility for users to do it, you know, when they want, without having to go in and, you know, swap everything out on day one and make it something new. So there's some extended life that they can still leverage with some of that older gear. And then I guess from Gen 3 on, we'll offer controller upgrades conceivably, keep your drives, swap out the controllers? That's the plan. the next several generations of technology to come. You know, how many years that is? You know, that's a to be determined, as I like to tell folks,

and it's probably a good place to end on. I've been with the organization for 36 years, and the only thing I can guarantee you is I'm not going to be here for another 36 years. So I would expect that, you know, when this architecture kind of hits the limits and it's ready for kind of that next gen, you know, maybe you and I'll be sitting on a beach somewhere,

having a margarita talking about, hey, the good old days. Remember the VNXE? I said you already told that story. Why don't you talk about the Gen 3 one? Maybe you'll have a good one for that. A better story for that. No, it was a good story. It was a good one.

All right. Well, this has been great. Again, you guys have been a gracious host. Let us kick around the lab, mostly unattended, which is dangerous, especially with our guys here. Oh, the security cameras everywhere, so, yeah. Yeah, but it might be too late for the habit. I guarantee there was havoc. Well, let's put it this way, if there's a latch broken on one of the nodes, we'll know who did it.

No, we won't know. We won't know. Okay. All right. Thanks again for doing this. Yeah, thanks a lot. Nice to chat with you guys. Thanks for the opportunity.

Good to have you. Great.