00:02
So we'll get to our introductions. Um, what's really cool about working here at Pure is I get to work with these great architects. uh, myself, I'm Zane Allen. I am on the principal technologist team, um, talk about strategy, talk to our customers, kind of where they're going to solve problems, and I'm joined by, like I said, my friends and colleagues here,
00:21
um, that I've known for quite some time. On our architect, uh, field solution architect, we'll start with you. Yeah, so my name is Robert Quimby or Q. I was on a team with a lot of robs and my focus is on everything Microsoft and soon Newtanics as we go GA with the uh co-engineered solution at the end of this year. Yeah, yeah.
00:43
You gave it away as in the presentation. Jeez, if you haven't heard that, you haven't heard the announcement yet. And then, uh, my friend Jerry Joe Hughes, how to y'all, I'm Joe, uh, legitimately around here. I'm great with a broom and I hold doors open. Um, no, I, I do a little bit of everything. I'm the flash that guy.
01:01
Uh, I was convinced to come over here to peer because I was the geek that said like, cool, I get to have my name and as an author for a Cisco validated design. So that's the kind of geek he is, yeah, right, that's awesome. No, uh, um, I'm big in the customer space for a lot of folks that know me, um, still lead 3 user groups, still very active in this stuff,
01:19
uh. Very passionate about this session in particular because I'm still actively a VMug leader, which was super awkward to have to leave my session, uh, yeah, VMware user group if if you know it's recovery there's a few of us left. But yeah, um, I do, I do everything basically, like I,
01:40
I built stuff in the data center for 20 years as a uh customer and then put auto automation around all of it. So, um, And you post on Reddit. Ah yes, so there's a session after this. It's basically this, right? So not, not a session. There is a community meet up, right, community meet up.
01:58
We, we are allowed to be outside the room. We are allowed to engage if you ask us direct questions, we are not allowed to drink all the free beer. We were reminded of that 3 times we cannot drink the free beer. They only told you that all the free beer, hey. No, but seriously, uh, yeah, there was, there was a random Reddit post that was put out,
02:16
sorry, I don't mean to call you a rando or anything, but there was just an ask for like, hey, what other customers are interested in doing a meetup? Maybe I can see if my AE can find me a room. Cody Hosterman naturally was the first one to say like I'm in, so lots of things started immediately happening, uh, and we sent it to our events team and Rob
02:33
Ludeman was like awesome, let's do this. And so within 3 hours I think of us roping him into that he gave our entire 120 person. Uh, staff call like an update of like, hey, we now have a session that's what, sorry, community meet up that is just this thing going on and no sales pitch, right? So, yeah, so, yeah, because it's it's perfect.
02:55
It was right after this one. So let's be kind of the precursor intro into that. I still wish we would have left you with $7 wordsmith. So that they used to say $7 wordsmith. uh, Joe is our emotional support FSA. Um, you were the, you were the hypervisor whisperer.
03:09
I don't know if you like that. Uh, so Robert has an interesting hack only one secret agent man is of Music, so I mean, yeah, it's yeah. What has happened to my deck? Come on. Do we break it? Oh, did it not hide all the things? All right. Man doing it live.
03:31
This is just a. Yeah, that is, that is not the, the beginning slide. I don't know what happened to our deck. So what we are, um, hoping to do here today, this is who's driving virtualization. So, um, we are, you gotta kinda come with us on this journey we're gonna take this kind of journey together down the road, um, to maybe alternate hypervisors,
03:51
right? That's kind of what the first slide said. Um, obviously, what we hope to do is we're good about this being interactive. We have the session afterwards, but it's really just kind of high level start to think of the things that one has to think about. Do I have the expertise on my staff? What is it really gonna take to move to this
04:08
other, uh, these other alternatives, you know, do I just stay with what I have today? Um, and this is where I think we've all started. Is anybody here not using VMware today? That was kind of in the, in the. OK, couple. OK, so you're thinking about going to VMware, or, OK,
04:25
don't do it, don't do it. Uh, so is that in the slide somewhere? I'm sorry, I, yeah, preview. It's awesome. All right, so, well, for most of us, this is where we probably started, right? This was the gold standard, the VMware, highway, you know, you can almost set.
04:42
You're right on cruise control you know the roads very well lets the organizations and, and you know, maybe you have the same experience, but um, let's the organizations kind of get, get, get the job done right now I have to think about the infrastructure stability, ecosystem. Many people are very familiar with VMware you could get the help,
05:01
right? The road was very smooth. That was VMware, right? But now broadcom owns VMware and there's a detour, right? I don't have to tell you new licensing changes, paradigm shift, um, there's detours in the road, right? There might be construction that's impeding your travel and you have to,
05:22
it's face a lot of organizations that we talked to have to think about, you know, hey, is there an alternate route I can take? Does it make sense to take an alternate route? What does an alternate route look like? Do I even have the map to that? So we're facing these detours and you know it's, it's a,
05:34
it's a real reality. In fact, I would venture a guess that everybody's kind of been thinking about this for a little bit and maybe we're getting to the end. I don't know. There's a renewal coming, right? And we really gotta, gotta make these decisions.
05:48
So what we have to do today is map that road together, right? So talk about the road so far, the many off ramps that we take alternate routes out there you can kind of see the theme we're gonna go to. We're gonna talk about some of the hypervisors and the things that, um, you might have to consider or you know the benefits, the trade-offs, so you can build your own map.
06:08
And These are the ones we're going to cover today, right? There's a few actually, there's many hypervisors out there that you could choose from. Um, there's many different off ramps you can take different paths that you can, you can make, and we, uh, our opinion some might be easier than others for you,
06:25
right? Um, some might make sense. I have the belief that some may find that multiple paths might make sense for different use cases, right? That might. Be a very viable thing for your organization and again it's gonna come down to that risk obviously that talent or I'm gonna say talent but that expertise you might have,
06:42
right, the type of drivers you have, do I have to retrain them? Like all, all of these things are gonna have to come into consideration, man, this is why we call him a $7 wordsmith. That's a long way to say it depends, you know. So yeah, and my slides are all out of order. I have to just go with it. So I don't know, I'm just trying to,
06:56
I mean. By the end of this we're hoping at least start thinking what kind of driver will you be, right? Will you leave the familiar? Will you take this new path? It's our opinion and this is kind of how we'll go through it. Um, we choose something here maybe on the left.
07:09
I think it might be the easiest routes to take, right, the most familiar, least familiar to some degree, a little bit of changes here and there, but still kind of the same experience that you have today, the same type of road, you know, maybe you gotta pay attention a little bit more, you can't put it on cruise.
07:22
Or you might sounds like at least somebody here might think about the hybrid cloud right where I'm really kind of changing how that drive looks right? I'm using a different type of management construct but might take me into new roads, new journeys, um, that I can for my management and then of course the completely customizable open source where I can probably customize it all but I might need the most expertise might
07:46
have the most uh change right the different um. Experience on on that path, right? KVM which is. Many things. So let's get started. So we have these little car themes. I actually didn't whip out the stickers. I made stickers for all these, but these are the only ones that came.
08:03
Um, mechanics AV. So it's kind of the revision of the classic car but maybe in a new way. I, I probably won't say that I think everybody knows what kind of car that is we had to blur out the logos, um, but you know, maybe a little bit more streamlined now, right? This is actually the the electric version of,
08:24
you know, the, the classic gas version. Uh, it's bundled with Newtanics. They put all the simple management into one. You still kind of get like a Venter management construct right again. So it's very familiar drive. It just, you know, the controls are set up a little bit different, uh, PRISM.
08:39
And if anybody's familiar with cars, the reason I have this here, uh, the way it is is just kind of a concept car on the road today because we're alluding to the thing that you already gave away the changes and the support that we're gonna have, uh, for it very, very soon, right, which is this here. Yeah, so with Newtanix, what we're doing is we're going to leverage compute only nodes in
09:03
Newtanics, and we did a video that we showed off, uh, last night and also at Cisco Live, and tomorrow we'll be repeating it at 4:30. A bunch of us are doing a bunch of live demos. Mine are mostly canned video demos because a lot of the stuff that I do, particularly in Hyper-V, uh, you click something and you wait 5 minutes.
09:22
So that's not a great interactive demo. So I've sped it up in in video. But what we're doing with these compute nodes is it's going to be a much smaller controller VM. That controller VM is going to connect to that flash array within VM of fabric TCP, so it's a different protocol. Newtanix doesn't have a fiber channel stack, and if you're familiar with how most HCI works
09:41
is that they own the device that controls the discs in the front of the server and so that was going to impede their GA date by a year. So we went to what's quickest to market and IAZ is painful. So we're gonna do this, uh, we're going to GA by the end of this year. All the management plan is going to be through PRISM.
09:59
So copy data management, right click a VM, snapshots occur on the flash array with a VA type RDM type model where every time you provision a disk in the TanX AHV that's connected to the flash array, it's gonna connect and create a pure volume, which means if you click on a VM and you want to clone 50 of them, boom, 3 seconds on, you know, instant on the flash array,
10:19
and then boom, you're booting the VMs. So a lot of cool stuff, a lot of stuff that we're working on, um, GA at the end of the year. And yeah, no, I'm glad you bought the snapshot stuff. So taking, right, so all the snapshots all of these are gonna be our snapshots, because that, that's another thing you gotta consider right when you're thinking about
10:38
changing hypervisors. What's, what's the migration look like? And I think you can probably speak to some of that, um. Uh, it's still Newtanix HP at the top. What's day two operations look like? A lot of backup vendors are supporting HV today. So again, kind of a very similar experience and, and doesn't obviously there's a conversion and
10:55
a change. You gotta figure out a new management plan, but with, with Pure storage kind of at the at the at the storage level you still get a lot of that that look and feel that you're used to. You know it's, it's really just this is the closest experience you're gonna have to Venter, especially if you've never seen any of these products before.
11:11
It's probably 80%. Uh, of what you're gonna get in VCenter because again like nobody chose poorly. VMware was the gold standard. They commoditized virtualization to where like your help desk could build things because they had right click functionality. You get most of that here, but, but also the same thing,
11:30
right, with this being new, like at release, it's gonna run with like their replication under the covers at first until we get ours in for a sync and safe to say closest thing to VCF. I'm not saying that, and we're gonna have a lot of automation if you want to convert to VA and then boom, you just need to migrate and move the OS disk, right? And since the VA is how this is gonna be a pure
11:56
volume, it's just manning, you know, instant. So same with our replication. If you're gonna set up active DR and just because you can't manage mounting the DR site in PRISM, well, here's a two line script. Do it boom, so you don't have to use Intanics, but it will work. So best of both worlds, so very excited about that.
12:16
I'm not gonna promise it's gonna be cheaper. Somebody else is gonna have to promise you that. And then there's Hyper-V. The cool is we have all the time in the world that clock hasn't moved on, um. So this is our um uh we debated a long time about these cars so you can I'd love feedback on it. It's the City Works truck, right?
12:36
It's that it's that utility drive truck. It's similar to, you know, it might not be able to jump in and drive there might be a little bigger might have different considerations controls might be in a different area, virtualization or, uh, you know, networking might be different, um, but it's something that probably already exists.
12:52
It might already be in the fleet. You might already have licensing, right? It's included, um, you might already even have the hardware that's there. And I think this is, yeah, we should invite a customer before we get we didn't introduce you in the beginning, but we kind of have a treat here.
13:06
We have a customer that, um, just recently, well recently came over to Pure and uh we were hoping, yeah, you gotta take Joe's mic. Awesome, glad everything's failing. So we, I'm gonna have Derek, I would like you to introduce yourself. Yeah, I'm Derek Vaughn. uh, I work for DR Horton,
13:25
uh, America's largest home builder since 2002. Number one, and what I wanted to do is just to spend, you know, 3 to 4 minutes chatting with Derek about his experience with Hyper-V part. Of my interest here is, um, he has done a lot of work with a lot of Hyper-V nodes. He's probably one of the top 5 customers with a number of Hyper-V nodes connected to Flash
13:47
array on the planet today. And some of the things that I wanted to ask is and has VMware in the environment and some VMware in the environment, but, uh, Derek, can you tell us a little bit about your experience with Hyper-V with our Active cluster, which is a synchronous replication? Yes, well, first off, it's been a great experience using Hyper-V with Active cluster.
14:08
Um, our Hyper-V clusters are stretched across two data centers, all on-prem. Uh, we have Excel 170s at both sites and our hosts run Windows Server 2022, uh, in core mode, and we manage that Hyper-V environment with System Center virtual Machine Manager. And the Hyper-V clustering seems to work really well with the active cluster technology.
14:36
um, it's given us the ability to really have active active data centers, which is something we've been working towards for a uh quite a while, which has been huge for us so. Now our Hyper-V hosts, we have paths to arrays at both data centers so we can essentially lose an entire array and still maintain storage connectivity
15:04
and keep the VMs online and that type of resilience has been a real game changer for DR Horton so. That's awesome. Can you tell us a little bit about a feature or some features on the flash array that kind of make your life easier? Yeah, I, I think, uh, the snapshots probably is, uh, saved us the most time,
15:27
my team's time anyway, uh, they're just fast, efficient, super simple to use, um, you know, we, we don't back up all of our non-production workloads, so having that safety net of that data that would otherwise be unrecoverable, uh, has been a huge for us also and just the ability to, to. To snapshot every volume that we have pretty
15:51
much instantly with no performance impact is uh is really, really huge for us um also you know we sometimes find that it's faster to do a VM recovery using the pure snapshot than it is with our traditional backup software uh hopefully not for long we're moving to rubric, but, um, and. Also It's Really just uh we have some
16:19
automation built out also with the snapshots so we can uh quickly snapshot production and put those down to the uh lower environments so that's also saved a lot of time for the team so yeah I'd say snapshots. That's awesome to hear and when you really just one more question um. Without, you know, getting into the weeds, can you at a high level talk about,
16:42
you know, the workloads that are running on the flash array, um, and whether the recent 18 months of change with Broadcom is impacted, you know, the decisions or your plans going forward because I know that you do have some VMware. Yes, we, yeah, we do maintain a small VMware environment, um, 400-ish VMs or so, uh, small compared to our Hyper-V environment.
17:06
So, uh, and, and the flash array itself, we run pretty much everything on the flash array. There's nothing that we can say no we can't run it on there unless it's, you know, like we're, we back up to flash blade now, so, um, but anyways, yeah, VMware we. Typically the workload that we have in VMware right now,
17:25
a lot of it, uh, comes from acquisitions. If we acquire a company that's VMware, uh, we just bring them over as VMware and we don't force an immediate platform switch at that time, uh, and also a lot of virtual appliances, uh, and third party software that really just requires you to run on VMware, our phone system being an example of that, um, but with the recent changes with VMware and
17:50
Broadcom. Uh, it does not really align with our long term strategy, so we will likely, uh, be migrating. A lot more of our VMware workload back to Hyper-V. Uh, we'll probably be using System Center virtual Machine Manager to do that, uh, VMware to Hyper-V conversion since they have some built in functionality with the newer
18:12
versions. So it's kind of built in tools, easy to migrate, yeah, it's just kind of conversion. Well, Derek, I really appreciate you coming up and giving your perspective and, and thank you, and we're gonna just go through a few more slides chatting about Hyper-V and everything. So thank you, sir.
18:31
So yeah, so obviously Nov Center, we talked about System Center, uh, uh, Joe's favorite PowerShell. Why nolo for PowerShell. Well, but, but also, if we're, if we're being honest about this, this is one of the biggest complaints that, that even all of us that are MVPs give the team at Microsoft all the time.
18:50
You have 6 different tools that all do a little bit of stuff. Not any one of them is like 100% coverage for all the things easily. Um, so yeah, it, it is, it is a pain point again, it's not, it's not, you know, single pain in the ass of BCenter. It's a splintered management experience, splintered management experience.
19:11
Wow, TMS. So, so one of the things that we do, right, we've got our PowerShell SDK, we have a backup SDK, ways to help you automate things, um, System Center, virtual Machine Manager, we can plug into that because you're gonna enable SMIS built into the array. You're not installing a custom provider with Windows admin Center,
19:28
we just shipped a brand new plug-in for the V2 gateway because they broke everything switching to .NET core, which was a good thing. Because you go from about 12 to 15 seconds per click for something to happen to sub 2 seconds and a lot of times sub 1 2nd, but it is a different experience and VMM they keep innovating and showing a new version of VMM. They just shipped last fall 2025.
19:52
But if you click 5 versions that go of VMM and compare it to VMM 2025, it's hard to discern any difference. And if you talk to Microsoft, they'll tell you that all their innovations going into Windows admin Center. So by the end of this year, you'll get a VM.
20:06
Migration tool and Windows Admin Center that's coming. Also a bunch of of the software defined networking and network hubs coming into whack. A bunch of things are coming into whack, and we're working on stuff with Windows Admin Center as well. So one of my calls for action or if there's functionality you would like to see in Pier's plugin to talk to your account team at Pure.
20:27
And open up an RFE because we did a hackathon we did right click of VM restore from Snapshot. It all works and they're like, Well, if we productize this, we need some customer demand. So help me with some customer demand on that. And if you need me to plant some seeds on like 10 ideas that we need RFPs for, I've got them, and we'll get we'll get our engineering to to
20:45
work on that. But yeah, management's changed. And we were gonna kind of talk about Windows Server 2025. A lot of changes in 2025, and if you wanna take this, um, uh, big, big changes just with Microsoft really focusing a lot more on on-prem data center rather than just Azure, um, believe it or not, you know,
21:04
Microsoft learned a lot of things by building data centers. It's one of those things that's a good byproduct. They, they have the ability now to actually scale a lot of this out into the environments with. A lot of the push for consolidating on single code base, really having support for Linux and the ecosystem,
21:22
uh, but they're trying to give that cloud management portal, right? It's all the tooling that you have from Azure to try and put it into your environment. Uh, Hot patch for VMs is coming, which. That's cool, you know, um, you still have to reboot the box at least once a quarter,
21:35
so it's not, it's not never rebooting the box. Uh, Azure Arc integration is also fantastic because you get, you get that similar experience to Windows Admin Center, right click functionality for a lot of things. A lot of the tooling that exists inside of Azure, even for plug-ins with,
21:50
uh, Sentinel, even plug-ins with a lot of Azure native services that you can now run against your on-prem workloads. Plus NV me over fabrics is coming. Sometimes And Uh, Linux, right. It, it'll, it'll support both, right? Um, I mean,
22:09
it's It's a lot of different things they're trying to attack all at once, but to be fair, they got a lot of people they actually went to their own teams, uh, that consume nothing but Windows or nothing but Linux and said, OK, those of you who manage our data centers at scale, what is your experience? What's your tooling? How do you do this?
22:28
And then basically said we need to go give that to all of our customers as well who don't have the internal engineering code for all of this. So they're getting there, right? Um, anybody who's played around with, you know, with current versions of Windows, either client or server, you know, you've started to see that there's even stuff like WG
22:45
or, or sorry, WGit, um, you actually have like true power shell-ish package management for Windows, which, yeah, it's very interesting, but they're trying to do a lot of things. They're trying to really, really marry the experience of what you have available to you on-prem. Match as closely as possible what you have available within Azure.
23:10
One of the other things that's fantastic about this is if you do have a lot of legacy workloads where your only option in the past was to move it to Azure so they would continue supporting you on, on ancient versions of Windows, um, they will eventually start rolling that in for Azure Local with Azure Arc, but at the same time, this is absolutely a subscription service that you pay like by the
23:31
feature, by the VM, by the, you know. But clock cycle, who knows that dovetails really nicely for Azure. Yeah, uh, yes, Azure Local is gonna be an interesting one. So, uh, we, we, we did our conversion van here so you can drive on Azure roads, but it can take you into those mountainous roads and bring that hybrid environment,
23:52
right? Um, you could say it sprints, sprints up the hills, plus, you know, we get to also poke fun at the whole cloud economics because. Why would anybody buy a, you know, cloud economics locally, right, yeah, expensive on prem. I'm just kidding, but it can be, you know, maybe, I mean.
24:11
Edge deployments, it could be a viable alternative. There, there were, there were a lot of disconnected products under the Azure Stack umbrella, um, that Tried to take off. They got a little bit of adoption and so they couldn't be completely abandoned. So they were spending a lot of engineering cycles on little side projects essentially, that just never took off.
24:33
They are now um realizing that as good as they are running data centers and as great as it is running Windows on-prem, they Honestly, you're just not in a lot of customer shops for having much influence other than just the fact that customers needed to run Windows based on an application workload. They're now trying to extend this out.
24:56
So, Azure Stack hub is here for now. It will, it will deprecate, I mean, probably in the next 2 years, I think. I forget what the, what the public timeline on that is. But Azure Stack Hub is still here, then you're gonna have, uh, Azure node scale. Essentially, this is the artist formerly known
25:13
as Azure Stack HCI. So this is the Windows flavoring that you're gonna run up to 16 servers. It's essentially server core, Hyper-V, right? Remote management for all of this. Then, We have rackscale, right? This was formerly Azure Stack operator Nexus.
25:29
This was what Azure custom built for Verizon and T-Mobile. They are now pulling the 5G components out of that to make it just essentially Azure Linux. That will run one from multi there is a storage vendor that supports it today. Really? Yeah, yeah, I wonder who. Yeah, we're, um, we're already validated for operator Nexus.
25:50
So as soon as this gets going, we will be there. Funny enough, the one that actually causes us problems is Node scale, which is Windows Server. So yeah, I, I remember you telling me that the team was amazed that he was able to like just PowerShell and make it work. They were surprised because they're not Windows people, right?
26:06
They're Azure local and they're focused. On the resource bridge that's installed that translates everything from the Azure calls into what Windows needs and they couldn't they were surprised that it worked, right? I'm like, guys, I'm using your stack, MPIO Microsoft Feature DSM, Microsoft Feature, Iai in the box at the hypervisor, Microsoft code, it all works on the flash array.
26:27
So the call to action on that is that this. Node scale Azure local Microsoft wants a couple of customers to kick the tires and not tell the storage experience but tell them the experience of all these Azure services and whether they can handle multiple storage paths, things of that nature. If you're interested uh in the next week we're
26:50
gonna be working with Microsoft on curating a list of about 5. Customs and we're gonna be rolling with this. So very, very soon you're going to see external storage support on that version of Azure Local. This version of Azure Local where we're already supported, has been rebranded Azure Local rack scale, and that's gonna have a huge announcement at Ignite,
27:12
and we're in private preview now with a handful of very large enterprises who are running this because this is already. Um, you mentioned some telcos, another big one at AT&T runs this. So if you have 5G, AT&T, anytime you're on that, you're running on this code for the last 5 years, 4 years on an X70. So this has been 7 9s of availability dial tone
27:35
service. Microsoft people this is so awesome, and they can scale to 128 node clusters. This one doesn't even run Windows. This is Azure Azure Linux, uh, but they have such high availability with this that they're rebranding it and sending it to the world. Hey, you want large clusters?
27:52
You want to manage it in Azure, boom, right there. So yeah, you're, you're welcome for all the AT&T customers until it doesn't work then not on that's not our problem. So, so definitely a shift you're gonna manage in the Azure world you're gonna manage it as Azure VMs, but, but a lot of benefits. Well, and,
28:08
and another distinction point too, uh, node scale, right? VMs only, rack scale VMs containers, at least at release, right? They're trying to have a clear delineation of like what they are intended for, so yeah. Yeah, so, yeah, containers and so forth. Awesome.
28:24
Data protection might be a little different, right? You gotta consider that might be agent based, but there are some definitely backup vendors out there that can that can work with it. It's still, it's still Windows, so like your on-premises tools will work. The reason that there was such a de-emphasis on any on-premises tools is that there's a feature
28:41
in Azure Arc and Azure Local where if you out of band, meaning you don't do it in the Azure portal, make a change, doesn't reflect up in Azure ever. So what they were doing is they're having all these customers doing Azure migrate from one Azure local cluster to another just so that it would reflect the change up in Azure,
28:59
which is ridiculous. So they called this a new feature that's coming called rehydration, and it's due by the end of this calendar year. So once that happens, all The stuff where you don't have all these backup tools you're having to use this Azure tool use anything that you normally use on premises. Instead of pointing to Hyper-V Cluster A, you're pointing to Hyper-V cluster Azure local
29:19
B and it'll all just work. And so we're working on a lot of integrations with Microsoft and that experience is going to be awesome. Sorry, I'm still confused. All right, I'm still confused by um the Azure local stuff, right? So like for the node scale.
29:38
Are you saying that this is just a Hyper-V cluster that you buy through Microsoft? Correct. It's, it's a re-scanned ISO that's Windows Server 2025 as of three weeks ago when it became Azure Local 2504. So the really, and the Azure integration with that part of it is. It it deploys a VM and it's called the Azure Arc Resource
30:08
Bridge, and that's different from Azure Arc enabling a Vanilla Hyper-V cluster. This is an AzureACo bridge managed in Azure and because it is Azure Local instead of Windows Server, in addition to all the licenses you're paying for Windows Server, you also have the privilege of sending Microsoft $10 a core a month. Yeah. And you're paying for the storage that you've
30:31
already bought it. No, so you buy that, yeah, that's you can scale out storage with pure storage storage is integrated. It's not consumed, uh, for like a bill or a charge or any sort of unless you want to buy it that way, we can do that too. But yeah, yeah, don't, don't confuse it even more.
30:51
Yeah, you're you're not, you're not paying for like bandwidth charges they're not, you know, egress to your storage that you own, no, yeah, yeah. So if you like that one, you're really gonna like this one then. Yeah, yeah. I think it's the more expensive one that's why I chose the G wagon.
31:08
Oh wait, I said that. So, uh, hybrid highway very similar. You're gonna get that cloud-like experience, but you know, the all wheel drive, you can take it off road, you can move into native services with AWS things like containers and such, um. Probably won't spend a lot of time on this.
31:25
We have really kind of announced this already, right? We also support it from a storage perspective. Uh, data protection is gonna be what AWS does, but we can, we can tie into that from the, from the storage part. But basically, at least how I understood it, and, and there actually is a how to guide
31:43
on AWS's site, you essentially have outpost, you buy pure you can put the two together and and scale out your storage that way. Yeah, we, we've supported Outpost for years. The real story here is that. That Amazon actually put in the engineering effort to actually make us an available option, um, so that when you integrate with Pure if you want just by default EC2 instances that you
32:03
deploy that are on outposts just run off the Pure array like you don't change anything. You don't have to pick a storage target or any of that. Yeah, it's co-engineered, it's integrated, uh, it's a lot simpler you don't kind of buy it together, you just scale it out. So if you like your cloud on-prem.
32:20
Piers got you. And then there's KVM. So I had a different EV up there. They told me I couldn't do it, so it was too recognizable but you guess what it was, um, but we did KVM everybody thinks it's this, right? It's a sleek, it's a new hotness, it's the edge device, right?
32:40
It's a sports car, if you will. I'm gonna argue that it's more like a commodity based import car that you've modified on your own because you have to put in the work, right? It's highly customizable, it's open source, right? We can. Do different things with it, but you're gonna have to know that expertise or you're going to
32:57
have to go find that expertise to manage it, right? migration to it. Day two operations, data protection, all those things. So there are some, some, some, some, um. Or this was our original idea. It's really like that old reliable pickup truck
33:14
that, you know, by the way, you know you've, you voided the warranty on the other one, right? There's no support, um, it's gonna get the job done or you can get it in different sizes, but you're gonna have to get under the hood. You're gonna have to do the fixing, right? It's not just out of the box. It's an MF and fold ranger, that's, you know.
33:30
But no, I like, I argued this one should absolutely be like that, you know, mid 90s like tan Corolla you see everywhere or that same like early 2000s silver Honda Accord. That's got to be the same one just traveling around. This is not to make fun of either one, right? That's a they are built, they are reliable.
33:46
They're absolutely battle tested, but you also see that KVM is everywhere, um, lot, a lot of, especially now in the last two years, surprisingly, there's been a lot of development of new hypervisor platforms, right? All these things are skinned KVM, right? That's this is all of them.
34:04
That's not true. The, the slide would be huge. Well, yeah, yeah, I mean KVM, like that that's the thing, and maybe this this audience knows that you don't move to KVM, right? KVM is just, it's in the kernel. It's I can do virtualization in Linux.
34:17
Um, there is no I moved to VMware. I moved to Hyper-V. It's kind of more direct. We just talked about all that. KVM is a different world. These aren't all KVM, by the way, uh, but things. Yeah, I knew somebody was gonna, yeah, but that's the point,
34:31
right? It all kind of gets lumped in here, um, and it's really gonna what you're gonna pick for certain use cases. Proxox somebody already called that out earlier it's probably one I hear about a lot, but I, I don't know where you all see it, but more in the slid, maybe more in the smaller, um, but it still gives you that kind of centralized management,
34:50
um, we ish. Uh, I go back and forth on if it has HA. We actually have another FSA on our staff that just released a blog on his, so I don't know I can share it. You can look for it, um, where he talks about how you can do with ice coy with Pure and actually get some siblings of, uh, HA, otherwise you gotta use something like NFS.
35:10
They're probably gonna be more for, you know, like smaller scale deployments where something like overt's gonna let you manage maybe larger scale environments, right, and give you more of that uh Venter experience. Um, a lot of this is also just that these are, these are all similar. Every one of them is different, right? If you look at even just specifically around
35:29
storage to, to, you know, think of what a lot of folks are gonna be trying to look at when they're moving to these from a VMware environment or from, you know, Hyper-V or anything else. Every one of them is gonna have their own caveats of what is supported and what is not. And if they do clustering, if they don't do clustering, if they layer their own file system
35:47
on top of stuff, how they deal with locks, how they deal with all that, and One of the big things that it comes down to for a lot of these is, you know, do I really care that this feature is supported by the hypervisor if it's supported by the underlying storage, right? There was a there was a question, uh, within a couple of days of us having that conversation about the meet up where somebody
36:08
said I'm moving to, um, XCPNG and like I just need to know. Should I do NFS or should I do ice cuzzy? And it was like, oh, totally ice cuzzy, you know, you do multipathing, it's totally built in the kernel, 100% ported. You lose thin provisioning and you lose snapshots.
36:22
Oh no, like we already do that, so rely on your storage to do it again, not the same exact experience. You're gonna have to like run like a PowerShell script to like wire it back up after the fact, but Yeah, it's a lot more manual, right? You're gonna have it, it is a lot more manageable. I mean, we, we found out,
36:37
uh, I think it was maybe 10 weeks ago now, uh, somebody had just released a plug-in for Proxox that they built on their own. Like I found out they requested access to a copy of the API docs, just built it. It's 90% ish correct. One of the workflows provisions data store that's read only,
36:55
which the vehicle plug in. Yeah, that, that was what they did. They went and looked like without having an array, without having access to the docks. They looked at a bunch of the blog posts that everybody had and a bunch of the old YouTube videos of us doing all the demos on the vSphere plug-in and said, oh, OK, so we think this is what most VMware administrators that have
37:15
external storage are going to do. And in the span of like 6 days, like wrote that up. Might be one of the great things if you want that level of flexibility inside of your environment to like make all of your stuff work. This is also one of the things that is very widely adopted in like university space or
37:34
anybody who is like grant funded. Uh, a lot of these things very much can be set and forget, right? So. Depends on what you need. There's things like scale, which really is like kind of an HCI, so kind of like VCF I guess. Um, and of course,
37:49
yeah, there's other type one hypervisors up there, Citrix, Power VM being the IBM one. I really like this. There's a BSD version called Beehive. I don't know if you're all aware of that. That's why I put it there. Because I wanna run BSD, so a lot of these are built for different use cases,
38:04
edge cases, maybe, um, and it really, it's all about finding the one that best fits your organization or again what I said earlier, maybe a multiple hypervisor strategy like our friends here at DR Horton adopted might work for certain workloads, but you gotta be careful in those considerations. And then OpenShift really kind of the containers of my virtualization,
38:26
virtualization of yeah, the reason this one looks at least significantly more polished is this one in a massive do that I had to work with Cisco on for a validated design and our marketing people were like, must say OpenShift. So the reason we actually put this in here is, believe it or not, one of the biggest advancements that sped up all of Microsoft, uh, getting Azure Local out the door.
38:47
Was for Azure Linux they just decided to actually adopt the sender driver, right? They're even now backporting this into server, not just Hyper-V. So every version of server, once they have sender connectivity in Windows Server, gets you direct access to an array, right?
39:06
All of your storage functionality will pass through to us. So the development we've been doing for 1213 years with uh sender for OpenStack. Now, immediately plugs in for everything as your local, right? We're, we're also hoping that more Alliance partners will do the smart thing and leave the storage to the storage people and,
39:27
uh, right, just use this shim layer for all the management. Um, but OpenShift is also an option, right? If you are looking for flexibility of like anything you want. It's there, um, quite a different journey than I heard it explained as an opinionated version. Oh it's, it's very kind of like this is the choices that'll make the most sense.
39:52
These are the choices you made in life but it is gonna be very, very different than what you're doing it, it is a fantastic offering if you need. All the bells and whistles that it has in it, but it is, it is very similar to VCF that it has a lot of things, right, um, integrated services and a lot of things that are very developer focused or very like
40:14
application, um, life cycle focused. So if like most of us, if we're being honest with ourselves, have 60% of their VM workloads that basically gets spun up, might have resources changed or increased at some point. They'll get tons of software updates, but they might just live there until they get turned off. Then you may not need a platform like this, right?
40:34
Run the thing that makes sense for your workloads that also, because we like our customers, doesn't have you spending all of your money all the time, right? And we have a ton of port works folks here at Pier spending a lot of cycles on this pretty sure they're all they heard us say all the words. So yeah, all it's all Ralph, it's all red hat.
40:54
Do me a favor if anybody comes to the community thing that we have next door when you go. See John Owings or just start screaming JO just say I wanna buy all the Kubernetes and yeah I will buy, seriously, I'll buy this whole room a round of beers like that and to tell him what Joe told you that he'll know doesn't support OpenShift anymore. He likes that one too. Yeah, so that was another really cool
41:21
survey that I don't get to show you. So who's driving virtualization? who's in here? You, you are the driver, right? You're, you're, it's the road that he travels up to you and Hocktan, yeah, Hockan. That's very funny, but yeah, I'll, I'll be nice and wait until the session's over.
41:41
I got, yeah, we had a, we had an awesome image of Hoctan actually by the same guy who wrote the, uh, multipath I scuzzy stuff who we acquired straight from Broadcom. So, um, us getting our new VMware field solutions architect actually cut the cloud solution architect for ABS team in half at VMware because that's what they're investigating. Yeah, we didn't even cover ABS up here,
42:03
but obviously that's VMware, but it's also something supported so yeah it's really gonna if you, if you need the easy offload of like I just wanna buy in for 5 years, I don't ever deal with Broadcom again. It's at least through Microsoft and like I want to get to the cloud and figure it out, but like I just need an answer right now I'm like. Maybe ABS is an easy button. It was a session right before,
42:22
by the way, yeah, so, so yeah, we would, we would have sent you to that one, but it already happened. Yeah, it's, it's definitely gonna be what your organization can do, what it's tooled for, right? Are you gonna stick with the expertise and the, the, uh, you know, the environment that you have today,
42:38
or are you gonna take advantage of kind of these modern. Modern ways to uh manage your VMs that, you know, might take you into that next 5 years.
