Podcast Archive - StorageReview.com - Podcast #111: HPE GreenLake Deep Dive

Episode Date: September 23, 2022

This week Brian takes us on an HPE GreenLake deep dive with Omer Asad.… The post Podcast #111: HPE GreenLake Deep Dive appeared first on StorageReview.com.

Hey everyone, Brian Beeler. Thanks for joining the Storage View podcast today. We've got a really good conversation coming up, I think, based on our conversation in June back at HPE Discover about everything that's going on within HPE GreenLake. topic, one that even as close as I am to technology didn't fully grasp in terms of the depth of the offering that HPE has. Omar Assad's with us today. We had a great conversation, like I said, back at HPE Discover, and it took me three months to pin this guy down to get this thing going again. So here we are, Omar. Thanks for doing this. Hey, thank you for inviting me, Brian.

Really appreciate it. I apologize for the three months. What I had told you at Discover, we actually had to implement and launch that as well in the same experience. You went and did something over the last quarter. Okay, good. No, I mean, you and I had a great conversation about GreenLake, but not just GreenLake, all the underlying componentry and what you guys have done there. And I think let's just start at a really high level because everything is as a service from the big IT vendors these days. It's an as-a-service lead first. And then, by the way, we've got some new hardware things underneath that are required to implement all these services. But really, services are the lead.

And HP GreenLake was the first to make that move in a big way, I think, and then really made a lot of the other big IT vendors react and come up with their own service solutions. And in some cases, your competitors announced concepts because they were, I think, pretty surprised by not just GreenLake, but the traction that you got out of the gate. So those are kind of two different things. So let's just kind of reset for the audience, if you will, on what does HPE mean when they talk about as a service and Green Lake services specifically? 100%. It's a very pertinent question, Brian. So when we talk about as a service, generally what the industry understands as a service is that I will continue to buy the same infrastructure and infrastructure solutions so as to be. But now it's a flexible consumption model, meaning I can lease this gear. There's a financial institution gets involved and the customer doesn't pay up front. They only pay for what they need. And typically that's what's

known as a service model. So you don't have to pay upfront and you can pay by the drip as you consume the equipment. And once you're done consuming, you can renew and continue your drip or you can change the solution. That's where most of the industry understands as a service. In offering the HPE infrastructure storage, networking, compute, and all the solutions as a service, HPE was kind of one of the first ones to start it. And then sort of everybody jumped into the infrastructure, it makes it a lot simpler. But then in order to go full all the way, you also have to make the consumption of technology simpler as well. And the biggest inspiration in that part of the world is actually two phenomenons that are happening in front of us. One is the SaaS way of

consuming software, that you go to some portal, use a SaaS service, and you don't have to do any heavy duty lifting. And the second way is the infrastructure as a service model from the hyperscalers. Because you can buy infrastructure as a service, meaning I bought storage, now working in compute, but I am paying by the drip, but I'm operating it myself, and all the complexity still remains there. Well, yeah. I mean, you're making a good differentiation, and I think this is an important one to understand. As a service, for some, is a finance model. But that only solves a piece of the problem, which I think is what you're about to get to.

What you're talking about is a fully managed service on top of the ability to flexibly consume these hardware resources. Exactly. So what is GreenLake? GreenLake is a cloud consumption model and a cloud operational model, meaning you can pay by the drip for as much as you use. But then in addition to that, how you consume that infrastructure is as a cloud operational model. The simplest and the closest example that I can give you is, imagine AWS Outposts or Azure Edge at scale. That you go to console.azure.com or you go to console.aws.com

and you log into that console and then you manage your infrastructure resources. In the case of AWS as one of the hyperscalers or Oracle or any hyperscaler, they are located in a region. And that's where those resources are. And you jump in through a SaaS console and you manage those resources and you pay for those resources. But if in place of that, that equipment was actually hosted within your own data centers or it was hosted within an MSP,

the GreenLake operational model and consumption model is exactly similar to that of a hyperscaler, that you pay by the drip and you use the GreenLake console to consume that. So cloud consumption and cloud operations. So simplify the procurement model. And other vendors have done that as well. Pure Storage, EMC, Dell, Cisco, they've all come along on that journey. But the next jump to be able to create a platform on top of which you can come in and then start to manage your resources

to give you exactly the cloud operational model that exists in the hyperscalers, but with the flexibility to be able to reproduce that within your own environment, that's what the GreenLake platform is all about. And that's why we had to build the platform that has the capability to register all this hardware and then abstract the day-to-day mundane, undifferentiated management of that from you so that you really focus on the workload.

So provisioning of storage, provisioning of compute, provisioning of networking, life cycling them, all of that becomes exactly like you were sitting on a hyperscale. So what this allows the customer to do is what I call an in place transformation of their data center. So let's talk about that because people that are going to this operational model are coming from, well, you tell me, I'm assuming that most are coming from on-prem data centers, but maybe some are coming from cloud heavy and pulling some of that back. I mean, what does that customer look like that's evaluating or transitioning some part of their infrastructure, either

on-prem or not not to GreenLake? A great question, right? So let's answer them in a sequence, right? So the first is a customer that is currently consuming hardware, right? Currently consuming storage, networking, and compute, maintaining their own data centers. Now, this customer can make two choices. This customer can say, hey, I can jump into the public cloud

and they should if it makes sense for them. I am not here to bash cloud. We are very good partners with AWS. We're amazing partners with Azure, because most of our offerings on the infrastructure layer are fully hybrid in nature. I mean, that's what you saw at Discover. Like the GreenLake console can manage your EC2 resources as well as it can manage your VMware resources on-prem, giving you a uniform operational view across the board, both rather than a fragmented view, right? So that's the first point.

So I really want the customer to decide what's better for their ROI. But then a lot of the customers have to go through this complex re-transformation journey. And what I call re-transformation is that they really have to simplify their operating model. In order to do that, rather than involving all of these professional services and ROI and cloud ROI and this and that, what we say is, start consuming your resources through the GreenLake platform. It is, your previous infrastructure is compatible with it. And you can buy new services through this platform as well. So it immediately gives you that cloud operational model, so that your installation, lifecycle management, patching, upgrade,

deployments, all of that become exactly like a hyperscaler. So your life is much more simpler. And then from the next point onwards, any new stuff that you procure that comes in on a pay as you go drip by drip basis. So your future looks extremely simple. Now you have your entire infrastructure simplified so that from an operational simplicity perspective, whether you are in the hyperscaler or whether you're on-prem, it's a level playing field. So from this point onwards, you really make a sound decision about, okay, I need to use these other platform services in the hyperscaler and it's really easy for me to get started. And then the customer can think about a migration journey. Because the GreenLake platform is fully plumbed into the hyperscaler of your

choice, starting with Azure and AWS, that mobility becomes really simple as well. And to actually really simplify the implementation of that, that is one of the fundamental reasons we were able to blend in a company called Zerto. We acquired Zerto about close to nine months ago. And it really allows for that simple click and point workload mobility between the hyperscaler and on-prem. So the first customer is that transforms in place and simplifies their operating model. And the second customer, yes, if a customer feels that their cloud operating cost is high, right, and they are the judge of that, right? We have different ROI calculators and they can use those. But if the customer decides that my AWS or my Azure or my hyperscaler cost is unbearable to me, what we tell that customer is that if you choose to move this workload back, the operating model will remain as simple. You're not going back to complexity you mentioned that's really important because as we look at all of these different things going on in the cloud and in the data center, we've been saying it, we, the industry has been saying it, it feels like for a decade that it's too complicated, that we're getting too many silos, we're getting too many things that are cordoned off and then need separate management planes. And it's getting worse again with all these AI

workloads, with GPU heavy infrastructure that now becomes a separate chunk. Or maybe I want to evaluate other options outside of the VMware hypervisor. Now I've got an OpenShift chunk or OpenStack or whatever, right? And we end up again with five or six different things. So having a control plane across as many things as possible, and I know you guys want to keep adding to that, which is great. One of the questions I have though, and you and I talked about this, is cloud consumption is one thing. And when you go buy AWS, you're buying a service level. You're not buying, and maybe you're buying some capacity or whatever, but you're not buying individual components.

You're not deciding, oh, am I getting this storage or that storage for the most part. When we look at the as-a-service offerings from the large infrastructure providers like HP and others, I often get concerned that all the hardware stuff that you guys have done for so long that is really a big differentiator when you look at the traditional, hey, do I want a storage array? Maybe I want a Nimble array.

Nimble's cool. It's got this info site thing. It's really easy to manage, easy to deploy. Those competitive advantages, I worry a little bit looking at the industry of those getting washed out when you go to this as a service model and you're saying, hey, business, don't worry about it. We're going to give you 53 terabytes of storage and whatever cores. And you have so much IP and skill and knowledge built up in these hardware systems.

What is that like as you try to maintain those differentiators, but also make it seamless on the GreenLake acquisition process? That's a very loaded question. And I'm glad you asked that question, right? So fundamentally, the first thing that we have done is, just like you and I talked about earlier, you said, hey, what is this GreenLake differentiation from as a service? So I said, pay by the drip, financial model innovation, but everybody has that. But then in addition to that,

the operating model, how you operate your infrastructure through the GreenLake platform has been dramatically simplified, keeping in mind that a hyperscalers console also achieves the same thing. And we really use that true as the North Star to give exactly the same experience to the GreenLake console on-prem. Now, the second question is, basing on that first as a service model, you're right. When somebody buys a storage device, they would say, hey, I like this storage platform. I would like to have it as a service, meaning finance release this from me.

That's what the world used to be. But now as we move forward, we go into service levels. So the way I describe a service level in the simplest way is that you have block service for block storage, you have object service for block storage, you have object service for object storage. When you buy this service, you buy this service with certain attributes. And those attributes are displayed from GreenLake as a rate card to the customer, right? What does that mean? I mean, I need these many IOPS, and I need this latency latency and I need this availability level. Availability level is six

nines of availability or in our case, being the only one in the industry to offer 100% availability guarantee with no strings attached, standard support contract. That's one of the different tiers. So the attributes that are listed on our wait card define the capabilities of what our platform has to offer to a customer. Because when it comes to IOPS and latency, everybody can do that. Pure storage operates at a sub millisecond latency all the way down to 150 microseconds. Elettra platforms from HPE operate all the way down to 150 microseconds. So latency and IOPS is not a differentiator in the industry anymore. That would be my first claim, right? Where the platform really comes

into the play is that how easy to lifecycle it is, how easy to upgrade is it? Does it work like a Lego building block and your storage moves seamlessly through the platform upgrades? And how are these things listed at attributes on your rate card is the key name of the game. So as product managers, what we have to do is keep your rate card simplified, yet differentiated. Right. So the way we have brought that to the market by saying, hey, you have a general purpose block storage service, six nines of availability, and you have a mission critical block storage service. The mission critical block storage service rate card says 100% availability guarantee, latency, tail latency fixed and consistent, IOPS up to this number committed to you. And that reflects on the pricing of the rate card. But once you buy or subscribe to a service,

then what hardware gets shipped to you, how it gets involved and how it appears as block service in your console for you to be able to operate it on a daily basis, that's completely abstracted from the customer. And that's where the simplicity of the offering comes through. So attributes of the rate card maintain the facets of the platform. Well, let's talk through some of that because if I'm in the cloud, in a traditional cloud provider, and I've got 100 VMs and I'm cruising along and 50 terabytes of storage, and now all of a sudden we have some success and I need another 25 terabytes of storage, I just say, okay, give it to me. And it's minutes, nay seconds to be operational. When you're delivering physical gear to a customer as part of the GreenLake service,

when you send that initial shipment, you're clearly over-provisioning a little bit in terms of storage and ports and cores and whatever. But how do you adjust to or how much do you over-provision to give yourself the headroom to expand with customers? How does that work? So this is where we are highly differentiated as well, right? So this is a bunch of intellectual property developed here as well. And the roots of it come from InfoSight, which was, again, the pioneer in the industry in proactive support management and proactive symptom management. We call, you know, once we deploy an array, we used to deploy an array at a customer site. And, you know, and as InfoSight sort of got absorbed into HPE, all of HPE's's platform jumped on top of it because everybody got the value of that. So we have had to do a lot of scale work in the background to be able to scale our InfoSight service.

And InfoSight service forms the basis of the proactive analytical service that sits inside GreenLake as well. I'm not going to use fancy words like machine learning and this and that and the other. It's nothing, you know, it's basically, we run a lot of deep inspection algorithms on the background to learn different patterns that informs our product management decisions. I didn't want to use words like digital transformation and AI power. Please don't. Right, because, no, we have four or five algorithms that we have internally developed. They crawl the data. They find out what our machines are doing, and they give us a prediction into the future,

and we feed that information to our product managers. Really, really, really simple. What we do have to do is we have to operate at scale because all our customers are sending us billions of data points. So there's a lot of innovation in terms of how we handle that scale. So then on top of that, Brian, what we did was we acquired another company called Cloud Physics. Cloud Physics has its roots in educating customers by observing their environment that would it suit better in AWS or would it suit better in GCP or would it shoot better in Azure or would it suit better in VMware? And then what

it does is it does, it has intellectual property to simulate your usage, learning your usage, and then simulate it into the future to give you a window of understanding of what more you need to buy. This investment was solely done to answer the exact question that you said. Look, at the end of the day, we have to deal with the physical supply chain. Suddenly a customer is bursting, what is happening, right? So in order to over provision, we use a combination even before we deploy a service at the customer. In our GreenLake console, customer has a cloud physics assessment that they can run. we use a combination even before we deploy a service at the customer in our green lake console customer has a cloud physics assessment that they can run which assesses their existing environment shows them their peak ratios maps them on the back end historical data that we have from other

customers in infosec and we give give a customer what we call a durability and an expandability score. And that informs our over-provisioning decision so that nothing is a surprise for the customer to be able to say, look, we have analyzed your environment. We have fished out your previous vCenter data because we know where to go and tap. We have observed your existing environment. And here are a few simulations which they can directly tweak in the dashboard to be able to say here is what the future looks like do you agree with it they're like yeah this looks good but I want to pad it further okay we pad it further and that padding is easy because they don't have to pay upfront for it yeah it's easy for the

customer right because they can tell you hey we think we might need another 20 terabytes of storage or whatever and you guys put it in there, and then you hope they consume it at a certain point. Right. And the beautiful thing about storage and compute is they always consume it. You give it to them and a service level that is desirable to them, right? The biggest shine in the picture is that suddenly their data center operations became as simple as a hyperscaler. But then at the same time, their availability and performance levels remain there. But they have constant access to this analysis service because we bought it and now we pass it on as free to our customers. So they can continue to simulate themselves in various different directions. Well, I would say, though, too, that when you,

from a storage first perspective, when you look at what you're consuming via GreenLake, you're still consuming real storage arrays. And not to say the cloud doesn't run on real storage, but it kind of doesn't. That's sort of the purpose of hyperscales in general. I know there's like ONTAP for FSX, and there's some stuff like that. And Dell's talked about cloudizing all of their software in Project Alpine. So there are things like that happening. But today, to be able to bring in your Electra arrays as part of GreenLake and still getting that deep wealth of data services

is still pretty important, I would think, to some of these customers that may not have had such a rich feature set from a storage perspective in the cloud. I mean, they still benefit from that, right? Yeah, I think that's spot on, right? Like you said, right? Yes, there is FSX, Cloud on Tap. There is Purity in the cloud. We have a few offerings that are there in the cloud marketplaces that customers can install.

And I'm not doing any bashing over here. It's just that the hyperscalers are very honest about their availability levels. I love that about them, right? Like Amazon, Azure, Oracle, GCP, this tape. Look, EBS service is five nines of availability. This is the level of durability. This is the level of tail latency that you get. And just, and I'm not forcing any words, right? I mean, you read those, they're very upfront about that. And when you compare that with what on-prem vendors can offer, there's a heck of a lot of a difference, right? I mean, on like cures and net apps and EMCs and HPs of the world, far superior performance

and far superior reliability and very rich set of data services. But the thing is, when you cloudify your offering, and now it runs on top of a combination, let's say an AWS of EC2 and EBS, the base durability and reliability is that of what a PBS is. So again, you've gotten a few more data services, but your durability and reliability has gone down and equidistant to the level of what the hyperscaler has to offer you. So when we looked at that, what we said was, look, a customer is running a mission-critical SAP application. The customer is running an Oracle application. They literally have to rewrite that application

in order to counter what EBS and other things have to offer for them in the hyperscaler. How about just bring that operating model to them? And that has been one of the roaring successes behind GreenLake as well, is what I call in-place conversion of your simplified operating and consumption, but walk me through the process of getting a little more about how this works. So I've got some worksheets and there's a little bit of consultancy there via software and sales team about what I need, what I think I need, what I really need. So we sort that out and you guys figure out how much proliance switching and storage that I need and put that together. Now, from there, how long does that take before that shows up at my loading dock? How does that physical bit work?

Great part. So the physical bit works differently in different services. The first thing that you have to recognize is just like you go to a hyperscaler console and subscribe to services, here because we are partner first, what you can do is you can sit down with your partner of choice or yourself, you can get an assessment using cloud physics, which is just like a SAS-based assessment. You click, assessor comes, and it starts to simulate your environment. It has all sorts of storage, compute, networking, VMware capabilities in there to be able to give you a good picture of what your existing environment looks like and what your future environment looks like. The beauty of paying the money for cloud physics is that it was,

it completely automated the assessment framework that, you know, you don't, you don't, you don't even need to be a data center administrator to be able to do that thing. You can, you can literally be in a CIO or, or a leader of an organization. It's click, download, install, run. That's it. And then graphs start popping up. So your partner has access to it. You have access to it, right? From that point onwards, now you have the capability to look at a service from HPE Greenlink. What does that mean?

Some services are software only, just like our backup and recovery service, or our disaster recovery service, which is built on top of Zerto. Some services are software only, just like our backup and recovery service, or our disaster recovery service, which is built on top of Zerto. Some services are a combination of hardware and software, just like AWS Outpost is a combination of hardware and software, right? So block service that you are talking about is a combination of hardware and software. Similarly, hybrid cloud service, the HCI service, is a combination of hardware and software. Block service is storage hardware and software. Hybrid cloud service is a full stack hardware and software, meaning it has compute, storage, and networking in a rack form factor built in together. But you buy it based on your CPU and memory, just like you would size a

particular VM. Once you have those two sizings done on top of the rate card, then you just subscribe and your channel partner puts in the subscription order. You get the subscription key and your account on Greenlight. And then the dock shows up in about 14 days. Stuff shows up in about 14 days on your docket. You wield that in your data center. In case of the hybrid cloud service, it's pre-wired. In the case of this block storage service, the equipment that shows up has three cables that are dangling out. The networking cable, the power cable, and the management console cable. You plug those in. you log into your console, you enter the subscription key, it automatically registers this gear with our fleet services. And then underneath the block service or underneath the

hybrid cloud service, this gear appears as the attributes that you want to start consuming. And there you go. It's literally that simple. And by the way, that has taken the most amount of time because the easiest thing was, here is my platform and here is a financing term. And we could have walked away. And that's what everybody else does. And that is the fundamental difference between GreenLake and what others do as a service. Yeah. So the fact that it comes in, the physical unit comes in and the customer racks it or their partner racks it or whatever, however it gets there. But you've already done all the setup, all the provisioning so that literally, and it knows its network IP, whatever, you've done all that work, so that when you plug it in, power it, it comes up, it gets its IP, it knows its identity,

and it's ready to roll. Right. And like for the hybrid cloud service, if you don't like the version of ESX that we have installed on it, it will ask you, hey, do you have a place where I can grab your ESX version? It'll grab it and fully automatically upgrade itself, right? So fleet service has, it will say, lifecycle this environment. It will not say this server, this ILO, this storage machine, this, that, and all of that is completely abstract. It is a lifecycle of this environment. And that environment doles out virtual machines. That's the service that you bought. Similarly, if you bought the block service, it is doling out lungs.

It is self-service. You can create multiple accounts for different people, just like you operate in a hyperscaler environment. It's just that the gear now sits in your data center and you're seeing the consumption levels on the side. So how does this change the ongoing maintenance? You talked about some of the lifecycle bits and that's really important in terms of,

I mean, all of the components, right? Especially on the servers, you've got drives, you've got NICs, you've got whatever, you've got all sorts of BMC, all the management stuff. That I know a lot of you can access and bubble up through ILO on the servers, on the storage, maybe it's a little bit different. But what changes in the maintenance of these things? So you talked about that the lifecycle stuff is there, but how much physical interaction is required from who would traditionally be an IT admin or a storage admin to handle updates that they would wait and do it Sunday at 2 a.m. or something like that?

Great point. Right. So when you have so when you have different services, if you have the block service or if you have the VM service or the hybrid cloud service. When you have the hybrid cloud service, everything is abstracted from you because you're just dealing with virtual machines upfront. But the intellectual property around InfoSight is that all our gear, whether it's the servers, it's the drives, it's the compute, it's the storage, it's the shelf, it's the DAC cabling, whatever. Everything is sending signatures to us.

So on a per customer basis, we have a state machine of what your footprint looks like and at what revision levels your footprint is. And we are constantly displaying that to you in a simplified form factor in your fleet manager. Your fleet manager in our SaaS console understands what different building blocks are operating in your environment, and we are constantly giving you advisory services on top of that. And by advisory services, meaning this is an update coming here, this is an update coming here. And all you have to do is you have to accept those things.

Once you accept those things on the backend, the software lifecycle management of BMC, ILO, firmware, drive firmware, decommissioning a drive, bringing a new drive into the system, all of that is completely automated. It is exactly how the hyperscalers do their hardware maintenance on the other side, because they have to do that at that particular scale. Now, if some capacity needs to be expanded, then we are showing you in the platform this has been shipped to you. Now it is sitting on your dock. Just roll it in. And as you roll it in, we automatically register and we add it into the mix of things, right? So from other than shipping of stuff

where you have to roll that in, the most amount of work with GreenLake was tenant management, customer management, multi-tenancy, security management, logging management, because this is an entire SaaS control plane that is now running your hardware on your behalf. And that's where most of the innovation has come in.

So in the hardware bit, do you have customers that don't even want that level of sort of de minimis hardware involvement? And in that case, how do they do that? Can they hire HPE to fully manage this thing so that I don't have to ever touch it? If a drive replacement comes, someone just goes and does it and I don't have to, basically, I don't, you know, not to put IT admins out of work, but maybe I don't want or don't have staff or don't want staff or whatever in these locations, especially, you know, we can talk a little bit more about small versions in a second, but if I've got some edge sites or whatever, what's going on there? I'm just sort of curious. And I think that's an excellent operational question. And I'm glad you've asked these

things because, you know, this is where most of the work went in because leasing would have been the easiest thing to do, right? The ownership experience of a cloud operational model is the hardest thing to do. So pertinent to the question, right? So the goal is, you see, doing day-to-day server maintenance is not the sexy thing to do, right? Everybody wants to do the most important work within the organization. But because we as vendors have not really paid attention to the operational simplicity

is where most of the complaints come in, right? Because if you look at the SaaS industry, you know, the biggest, why does everybody love Workday? And even IT loves Workday, right? Because if you look at the SaaS industry, you know, the biggest, why does everybody love Workday? And even IT loves Workday, right? Nobody wants to maintain an HR and management system with patching and this and that and the other. It's like, dude, yeah, we have subscribed to Workday. Please use it, right?

And HR guys love it because it's just a SaaS console. It's so simple to use. And the IT guys love it because they can go do other productive stuff. So SaaS introduced a level of simplicity in software management and software rollout and software upgrades that now the back-end vendor takes care of. Similarly, doing infrastructure as a service on-premise is the next level of innovation from GreenLake. So there are two models. So you can get the gear as services, and then you can operate it through

our console. And we're doing a lot of the work in the background. And you can rack, stack, change drives yourself. But entire software management is completely, lifecycle management of the software and the hardware is completely automated for you through alert management. And we have written streaming systems to be able to collect this data and then instruct them back to the customers. Those are major components of the GreenLake SaaS platform. But then you have the next level available as well, because if you really want to be hands-off, what you can do is you can create an account inside because it's a completely self-service account. You can create another account for HPE's professional services teams to log in and manage your systems on your behalf. And they will mobilize the physical staff at the back end to jump into your data centers if you give us the badges to be able to lifecycle drives and put machines in and all of

that and that entire system is also fully controlled by you are fully you because you are the owners of the console you're fully looking at what's going on but the operator rather than your IT administrator is actually a professional sitting at HPE's support shops and because you have allowed them in by creating an account for themselves in over there. And that's another level of service that you can subscribe to. Yeah, I mean, that seems like an interesting modality for distributed enterprise retail. I mean, a lot of these use cases. And actually, before we even go down that road, I want to talk about sizing, the size of GreenLake.

Because I think when organizations think about these new models, these as-a-service models, that a lot of mid-size and smaller businesses think, well, that's a Fortune 500 thing. That's not for me. That's multi-rack, petabytes of scale kind of stuff. That's not for me. And I do think we see a lot of smaller orgs, even midsize, sort of opt themselves out of as a service before they even really consider it. And when you and I talked, I think if I recall, you talked about configurations that were as little as for that hybrid cloud offering, two servers, a switch, and a little Electra,

like very, very small. Yeah, yeah. And those mid-market customers can just order the hybrid cloud service, click, next hybrid cloud service comes, meaning rack showed up, fully configured with VMware. You can literally remotion your workloads on top of it and you can start doling out virtual machines.

No ESX management needed, no storage management needed, no network management needed, no disaster recovery management needed as well because this thing can peer to a hyperscaler and Zerto can start to mobilize your workloads back and forth, right? So all of this heavy lifting just completely went away. Backup as a service is built into that. So it's automatically backing itself up and it's dumping into a repository if you have one, or it is pushing that into a hyperscaler for your offsite backups as well. Point in time recovery is available. So you just focus on your workload. You can order a system saying, hey, I need a system for 30 virtual machines at this performance level. That's it. And you order the hybrid cloud service and it shows up. You can, from a menu of options, say DR enabled, the backup enabled, done. Forget about it. Okay. But what about the price though? Because that's the question. Because when I go and look at a, if I go get an entry array for

20, 30 grand, a couple of servers, I don't know, a switch, maybe I'm in 75, 125 grand. If I'm in $75,000, $125,000 if I'm going to go buy a real small infrastructure. I don't want to sign up with you for three years and pay double just to simplify my life. You know what I mean? So what's the balance there or how small can you get? Do you guys have any public or rate card that you're willing to talk about on how small these things go and what the affordability is like? Yeah. how small these things go and what the affordability is like? Yeah, so rate cards are available on partner portals because different, we don't speak about the pricing openly because different partners enjoy different levels of discount with us based on HP Platinum partners, HP Silver partners, right? What I can tell you off the bat is that it is a 40% improvement on your ROI efficiency right off the bat.

Because, see, you have DR. So when you subscribe to a hybrid cloud service, HCI service, you can subscribe to just a virtual machine as a service. Like I want one VM, $4 a gig or whatever. But in addition to that, that backup, which is included into it, the disaster recovery that is included into it, that just takes it to the next level. But the thing is,

it depends on the customer's adoption timeline. Because a lot of the customers have existing backup solutions. They have existing disaster recovery solutions. So they want to draw them down completely before 100% switching over. So what really matters in this adoption is at what part of a journey that you're signing up for. But what I can tell you with a lot of assurity is that if you are a green customer just starting on a brand new workload to be able to order a service which has disaster recovery backup virtual machine capabilities all of that included in it it's an instant gratification start for you where you

don't have to do any of those other things. So yes, there is a price of convenience, but from an ROI perspective, you will always be cheaper as compared to just investing everything upfront. It's fair, but I was happy to have that part of the conversation with you because really, I think in my mind, maybe I just sort of lulled into my own belief that this is a big company problem. And not necessarily the case, I think, is the takeaway.

So I was asking, too, about retail and distributed enterprise and multi-location. Is GreenLake a fit there, too? Or how does that work as you look at, you know, just pick retail because that's easy. I'm in Cincinnati, so I'm always thinking about Kroger and Macy's and that kind of thing. If I need infrastructure at a couple hundred or thousand locations, does GreenLake fit for that? 100%. I mean, public reference, Home Depot, right? I mean, they're everywhere, right? They're a flagship product for this. And that flagship product is actually all the way, I mean, from GreenLake fits not just on the compute environment or the store's execution environment, but all,

you know, because of the flexibility of having Aruba into the mix, that literally branch as a service concept comes in for them. Because they can pick and choose multiple services to make a bigger construct that sort of takes care of that entire branch. But if you're just looking at it from a compute or like a, you know, what I call a hybrid cloud service, it can go as low as the physical asset that shows up is like a two-use server that has storage networking compute in there, like a hyper-converged form factor, but you are paying on a per VM basis. And then it can go as large as 64 compute servers, two petabyte array, massive 40 gig top of the rack switches. But the beauty of it is you did not have to choose and configure

any of this hardware. Our machine, our translation engines did that at the back end based on the rate cards requirement that you selected. Sure. And your decision was informed

by an algorithmic simulation that ran in your environment for as long as you wanted to be able to feed into that engine. And then we say, yep, if your application requirements are this much storage, this much IO, this, this, this, this is a different minimum commit levels that you can start off and the right gear then starts to show up. Yeah, that's pretty remarkable.

Hopefully, if I'm running gas gas stations i don't do it wrong and get the 64 node rack showing up there's only so much room for cigarettes and servers in those little booths right uh but yeah i mean that that that's pretty cool i mean it's uh it's it's really fun to think about the possibilities i think operationally operationally more than any. So I know it's not new, but it's in this incarnation is kind of new because you've got all sorts of new stuff that they've been folding in to GreenLake. So what's some of the feedback? Like Home Depot, if they're referenceable, when they got fully realized with GreenLake, what benefits did they find or did some other customer find that they weren't expecting? Because I think we all expect sort of the ease of operations,

but there's got to be some other nugget there that popped up that maybe wasn't expected. Planning becomes extremely simple. Planning becomes extremely simple. Operational planning becomes extremely simple. Because they're not worried about having to buy another whatever. It's like, yeah, hey, if I need it, I will be informed about it. Or if somebody has a need coming up, they can burst. And then that need can go away. So the burst can come down. The procurement cycle becomes a lot more simpler, right?

Once we are standardized on a deployment methodology, it just becomes rinse, lather, repeat, and all of those attributes can be saved in the GreenLake console, right? The way, Brian, the simplest way to understand this is I use this term with customers, which is called in-place transformation. You can transform by lifting and shifting to the public cloud because you like that operational model. But then using the GreenLake console, your burstiness into the cloud

or mobility of certain workloads into the public cloud also becomes really, really simple because the disaster recovery and the backup and recovery capabilities that we have to offer fit in that customer's journey. So, and going back to the first point, right, the very fundamental point, financial simplicity came in, but that is there with every other vendor. That's not the differentiating part. The operational simplicity, the day two to the day 360 operation, that operational simplicity, that predictability,

the ability to literally take your block storage device and converting them into an internal block storage service is extremely enlightening to customers that okay in an organization I am a storage admin but my staff doesn't have self-service capability I'm always the bottleneck so the cloud operational model of the block service from HPE means that I can now create an account for Brian, for George, for Tom, for Omer, and I can define their playing grounds and they can self-service themselves. Similarly, on the hybrid cloud service, I can increase the scope of their self-service all

the way to the VMs. I can ask them to take a snapshot of a running application and then reproduce that into their own environment through the VM as a service. So writing all of this automation is what took two years and then bringing that console back out to the customer. Leasing, we've been doing leasing as a service like everybody else for the past four years. Yeah.

Okay. Well, I mean, the two years of work to get here is great, but that's baked in now, so no one cares about how long that took. You've got to go forward. And so I know you can't disclose a lot, but as you think about what comes next, where do you have to keep innovating with green lake i mean we talked at the very top about more and more silos showing up with different infrastructures red hat stuff gpu stuff other accelerator whatever like what what do you have to do to keep being open to what's next so that you can keep folding in these additional

services so the first thing is you know uh it's it's a it additional services? So the first thing is, you know, it's a journey, right? The first thing was, yeah, everybody has the first thing called financing, right? Let's simplify the procurement model. We sort of took the bold step of simplifying the operational model as well, right? So you have the on-prem hyperscaler team going on, right? That's what's there. The next step after that, so the first step after simplification was abstracting the hardware, right? So we abstracted the hardware and really

brought true infrastructure as a service through a cloud operation and a cloud consumption model on-prem. The next step is to jump at what I call application frameworks, in my opinion, right? What are the common on-premise applications that require a lot of complexity to set up, to troubleshoot, so that those applications can then be abstracted into as a service models as well. That's our next goal going on. You know, container as a service,m as a service we got that for free

because we abstracted the infrastructure so so the exposure was at a vmware level or a container level that's done right so hpe private cloud which is offered with green lake as a service hpe hybrid cloud which is offered with greenLake as a service, block service, networking service, which is called NAAS, network as a service. And then on top of it, backup and recovery service, Zerto, which brings in disaster recovery as a service. These are individual services. The next step is to bring these services as application as a service offerings so that the customer doesn't have to deal with that as well, right? And then keep jumping one step on top of that, one step on top of that, so that the consumption and operation at the data center becomes as simple as we can get it down to.

So then the decision between a hyperscaler and onprem, or an MSP literally is an accounting decision, not a complexity decision of owning the hardware. Well, and I suppose too, though, one of the benefits to this model is that as you roll these things out, they should be things that you could just pick up and consume in your console, right? I mean, it's not, with some exceptions, I would guess, it's not hardware dependent, right, to enable, you know, a lot of the, it's not hardware dependent, right? To enable, you know, a lot of the, especially the software enhancements, right? It just comes along with your Greenlight. Yep. A hundred percent. You got it. All right. Well, this is cool. You know, I think it's great to see where this is going. You know, I think the model's really interesting

and your delivery mechanism and the fact that you can scale down to really small is also super beneficial. So I'm enthusiastic to see where this goes. Where do we send people? HPE, GreenLake or some such to learn more about this? GreenLake.hpe.com. That's the best place to come and learn about it. Or a partner portal for all the partners that are listening out there, jumping on the partner portal, getting access to the console. I encourage as many people in the partner portal for all the partners that are listening out there, jumping on the partner portal, getting access to the console. I encourage as many people in the partner community, in the customer community to jump on the console and see what the console can. And console is also the best way to learn about these services, right?

We can, obviously, there's the marketing and the demand engine that works in. But then again, I'm very thankful to you as well, because, you know, one of the very few people that ask operationally detailed questions. So that really brings out the what we have built. So I'm really thankful for that as well. Well, we're very smart over here. Absolutely. Absolutely. That's the key. No, but I think it just comes down to having these conversations and you've got to be willing to kind of invest the time. And even for you to drop an hour or whatever it's been into this podcast to sit here and field these questions. I mean, most of the times when we go through the briefings, when you launch new products or whatever, it's 22 and a half minutes to kind of burn through slides and whatever. We can never get real deep.

That's why I love these conversations so much because we can, without much time pressure, really get into it and really sort of drift through all the different angles and where the benefits sit. So this is a good format for that. 100%. And thank you for taking the time. I really appreciate it. I'm glad you did it and I look forward to seeing you

hopefully again sometime before next June. 100%. All right. Appreciate you doing the pod. Thanks so much. Thanks, man. Appreciate it.