What’s New in VMware Cloud Foundation 9.0

Hello everybody. Thank you for joining this session. Uh, obviously we're out with VCF nine.

This will be a fantastic highlight of what's new, what you should know, how you can start playing with this fantastic platform. We have a bunch of folks here on this session today. We're gonna make it interactive.

We're gonna go through a bunch of content. We're gonna have about 30 minutes of, let's see what's new. Let's see how you can te take it for a drive.

Let's see what's changed in the past, let's say two years of Broadcom and VMware coming together to work on this. And then we're gonna have a reasonably live conversation about some of the stuff that you're gonna see as part of these slides. There are other sessions coming after this.

Make sure you check in. Some of them are gonna go far deeper than we're gonna go in the next 30 minutes, so tune in to what comes after this. My name is Sabina Anya.

I'm your senior technical product lead for VMware Cloud Foundation. Been with VMware for five years now every time VCF. Let's have a little bit of fun today.

Cloud Foundation has been around for, I don't know, a good part of seven to eight years, and we've had vSphere for about 25 years. Don't count the years if you don't wanna look at a calendar. The fun part of this is, is that at the core of what VMware is, it's always been about virtualization, taking hardware and making the most of it, allocating the resources in the best and most efficient way.

But why stop just at virtualizing servers? So for the past 7, 8, 9 years, depending how you're counting the the age of VCF, we've basically looked at virtualizing storage, networking, and tying it into our automation and operation tooling that we've had in the past and basically creating this one platform for everything. The last version of VCF before nine oh was five two.

Up until that stage you could still see like the seams between all these different components and it came more of a, you know, use what you want. But ever since we kind of joined forces with Broadcom on this, one of the major shifts that we've seen and we've executed on successfully is this, it's one platform. You don't have to start trying to understand each separate product anymore.

We're delivering to you one platform, VMware Cloud Foundation, that that hosts both containers and virtual machines that has automation at the fingertips and helps you do day-to-day operations. And this can be deployed, you know, on premises in your, you know, Paris, London data center. You can put it in a partner cloud, you can work with some of our many partners to do this, or you can put it in smaller remote locations and have a smaller footprint of this.

All of this is your VCF fleet now that's supposed to deliver on your solution, your technology and your applications that you want delivered Quick, kind of what does this do for you? Why should you care? Again, we're gonna get technical in a minute, so bear with me as we go through a little bit through why should you the what is, sorry, the why should you care part.

First of all, when you have one platform, it's a lot easier for you, for your business leaders to increase revenue quickly, changing market conditions, your management wants you to start thinking about how to deploy ai. Do you wanna really reinvent your data center over the weekend or do you wanna say, sure, we support this out of the box, we support containers out of the box, we support whatever you need next. We probably thought of it and put it in the platform, so why not just take advantage of that?

Also, because we've introduced so much intelligence in the platform, you're able to look at different issues across the different, uh, technologies, whether it's networking and storage, and combine that with, again, our automation operations platform and mitigate problems before they escalate. Do some application troubleshooting. We actually have a little bit of this to show you later.

Again, downtime is just not a word that's okay anymore in 2025, it wasn't okay in 95, but there is zero tolerance today for that. So being able to minimize, to recover fast, to be secure and resilient becomes key to all of this. And last but not least, there is not one IT outfit out there that does not want to lower cost.

There is not one IT manager that suggests, Hey, you know what? We should have this year more cost and less efficiency in our, in our it. So part of what makes a platform great is that you can look at your entire footprint.

You have that level of visibility so you can reduce waste. How many times have you not deployed oversized VMs? Because you know, nobody actually gave you any guidelines on how to do this.

And it's like nobody got fired for an oversized vm, but now you have all of this space being eaten up all over the place and your management is saying, can we do more with less? This is one of the places where working with a platform that has utilization and cost insight can help you build a report in minutes rather than days. So what's changed?

Most of us are used to doing IT infrastructure by basically managing each and every component in our IT zoo. Whether it's I'm managing my vSphere, I'm managing my whatever storage environment I have my entire data center, lan, so on and so forth. We don't think of the IT environment as one continuous delivery platform.

And sadly, it's also how we look at different failures across the data center. We bounce tickets around between the different teams. However, VCF nine absolutely wants to rewrite the way we do this.

And we're very adamant about it because you have the traditional sort of the bottom layer of the, the technologies that you know and love with vSphere, storage, powered by vsan, networking, powered by VCF networking that bring together, again, once you deploy VCFA, VM has everything IT needs. It has networking attached to it, it has storage attached to it, everything you could need to make your workloads run. And then we have all these adjacent services.

You can do that capacity management dimension, you can do configuration, you can do multi-tenancy, you can do fleet management, upgrade your entire data center kind of in one flow. But all of this is table stakes or IT AB absolutely should be table stakes at this point. What's on, what's next?

What goes on top of this? The way we consume infrastructure cannot be manually creating VLANs and clicking through screens in order to deploy VMs. It has to be design ones repeat every time, or at least modify little by little what you need and kind of get it to that state where you can design once and deploy every time.

So having something like an easy to use service catalog, having orchestration that doesn't take two years to set up, having the ability to have a content library that's easy to access, whether you're talking about infrastructure service deployment, or application deployment, all of this is part of what VCF nine offers. And you can take it to the next level. All of this is by default, PC nine.

You wanna enhance that experience? Let's do it. Let's look at things like private ai.

There's a lot of buzz in the market now about using the AI to either in pharma to analyze data faster. Again, work with all sorts of different workloads. LLMs are definitely relevant.

How can you deploy that on top of the data center? There are so many dependencies when you deploy ai. It's becoming increasingly difficult for IT administrators to tackle all of the challenges.

This is what the advanced service does for you. Other ran services include security, ransomware, recovery load balancing, all sorts of things that you can enhance your VCF experience with. But get the basics right.

And let's talk about what made VCF nine this fantastic platform that you need to know about. This is the, the crux of what's new. First of all, you get one interface to private cloud operations.

So think about one console where you can do day-to-day operations, where you can manage storage, networking, all of the things that, uh, attached to that. And then when we move into that consumption layer where we wanna define services catalogs and so on, you have one interface for that as well for all of automation. And for those of you that are familiar with Aria, that's not the experience we're offering you.

We've simplified, we've made it easier to consume. We've made it a lot easier to actually configure your flows. So it's not the VC we used to know.

The next part is we wanted you to have feature purity when it comes to running VMs and containers to the point where you don't have to think twice about which kind of workload do I need to deploy today. We've had container support for a very long time. We have something called VKS that's part of the platform vSphere you known and loved again for about 25 years.

So between these two, VCF nine is in this perfect place to deliver all of these out of the box without any additional add-ons or without you needing to go shopping around for additional technologies. Second, we have a, sorry, fourth item on the list. We have a lot of security built in straight into the platform.

You can enhance this with any of our add-ons, but the most important part is it is built with security in mind. It's not a bolt-on. Every single technology in there has a security work stream attached to it our engineers are working on and constantly refining it.

It also allows you to build your own sovereign private cloud. Last one, remember how I mentioned that there is no IT leader that wants things to be expensive? We've brought cost control front and center in the platform.

It's becoming easier and easier for you to track your consumption and your cost. And I'm gonna show you how that looks like in the platform. Just a bit of a highlight on, you know, what have we done for innovation and efficiency?

We've talked with a lot of customers before and after we've launched VCF nine and some of the things that they cared about was definitely, it's great that we've delivered a platform. It's great that we've kind of removed the seams between our different technologies, but have you forgotten about your core platform? Have you forgotten kind of to really, really innovate and bring out all of those cool features that made VMware so so incredibly sexy?

And no, we haven't, we've worked on our memory tiering with MVME. If you're unfamiliar with this, think about it like this. You have hot pages that need to stay in memory, your expensive memory, but you also have cold pages with low rates of access.

At which point why are those staying in your incredibly expensive memory? Why not downgrade that to something like MVME that can perfectly support the rate of access to that? That means at the end of the day, you get about 40% reduction in server TCO.

And we've done this again, across different tests, across different workloads such as databases, and we've seen that almost consistently you end up needing 40%, sorry, you reduce 40% of your server TCO. So Sabina, with that feature, um, I know we're probably gonna get a deeper dive in other sessions, but how is memory tiering different than traditional VM swap for people who are familiar with that feature? This is largely how the page allocation is being done in memory as opposed to how we've done then traditionally.

So you're thinking about you had to allocate all of the necessary memory for, um, the different workloads and how they load this into memory. Once you have all of these have open pages and what the MV metering algorithm looks at is the rate of access and the rate of open across different, sorry, the rate of load across these different pages. How hot are they?

How cold are they? And it basically shifts anything that it doesn't need or it doesn't have the above a certain threshold rate of access onto the NVME drive. So you don't need to, you don't need to provision as much.

You can run it over subscription algorithm, something that you couldn't have done before. Gotcha. Okay.

On the second feature that we wanted to bring up here is the 34% lower storage TCO with global ddu. Key part here is we know that global DUP isn't necessarily that new in the market. There's any number of other storage system that had this before.

However, VS A-N-E-S-A has brought a lot of improvements. Global DDU was just a natural next step. What's interesting about this is, and again, if you're unfamiliar with how Global ddu works, we basically eliminate duplicate data across the cluster in real time.

You need fewer drives, lower power and cooling and everything that's associated with when basically you have a higher storage footprint. As a result of this removal of the duplication, uh, the duplicated data, you end up with 34% lower TCO. The switching performance is an interesting one because we don't own a hardware layer per se.

We, we have virtualized switching and raving. What's really interesting about the NSX enhanced data path is that it gives you the ability to leverage the full, um, transport pipeline that we have. So at the kernel level, once you start loading data, you need to be able to process it fast enough in order to be able to take advantage of line rates.

Switching. A lot of how fast we process the package becomes incredibly important because if we can process the package fast enough from our networking perspective, it doesn't matter if you have 40 gig or a hundred gig on your networking side. So what we ended up doing here is make it a lot more efficient to pass packets through our own pipeline once they get out.

Again, that's what the mercy of your switching layer. So we ended up improving it three times, coming to near line rate for the switching environment. The next one is compared to bare metal.

There's an, there's an interesting statistic here. So we've done some ML performance benchmarks and we saw that VCF retains nearly 99% of bare metal inference performance. Now think about it like this, when you virtualize full bare metal, the tolerable amount of performance overhead is about nine to 10%.

That's the general standard for this. We managed to get it down to 1% as a result of this. So you get all the benefits of virtualization, the mobility, hint, hint, vMotion, the automation, the security without sacrificing GPU throughput.

And all of this, by the way, you can read, there's an entire report on it. And the last bit that we kind of brought in the core platform is this vMotion for ai. So we have vMotion for I think 18 years if I'm to count my years, right?

The very interesting part here is that we didn't have it for VGPU. Now if you have workloads running on Vgpu, we have this vMotion optimization in place where you can do a live vMotion, even for large AI ML jobs. This means, again, a truly zero downtime environment regardless of the workload type.

That's a lot of the, what our core platform is doing. If you're in a position to think like, okay, what can I do with VCF? Clearly this is a massive platform.

Clearly there's a lot of technology in there. How can I get started? Like, what should I look at first?

We have a few basic use cases that we'd like you to look at and we've split them up in these three different, uh, pillars. First of all, we have infrastructure modernization. Basically, if you're coming from legacy environments, what would you like to think of first?

For example, maybe deploy private cloud infrastructure on premises. Like how do you do that? Like that, that day zero.

How do I get started? How do I put together a green field that doesn't take months? The second one is optimizing infrastructure and operations.

Again, if you don't want to keep bouncing between your different environments to troubleshoot, to set things up, to deploy VMs and containers, how can you optimize this better? And then there's a third step. Where do you want to extend this?

Most customers have a hybrid cloud strategy. Nobody's entirely on-prem. And what we wanna do is help you go to the edge to sovereign clouds or even in the public cloud.

Application modernization means have your data center, have your private cloud, have your VCF deployment, be ready to do self-service for different cloud services and for app teams, just the, the exact same way that you deploy services in the public cloud. Like for example, you request a VPC in the public cloud or you request a slice of storage the same way you can actually make that available in VCF nine. Are you Kubernetes ready today or you'd like to be, again, I've mentioned this a few times before.

You have Kubernetes straight out of the platform, and if you wanna go private AI, that also works. Security modernization is a key topic as well. Compliance and hardened infrastructure has been something that's been difficult for companies for a very long time.

Compliance has stronger and stronger rules every year, and infrastructure is sprawling out every year some more. So it's constantly a catchup game, but you also have things like ladder security, intrusion prevention and detection that you want to look at. Last but not least, ransomware recovery, one of the hottest topics in data protection.

How can you do this with VCF? Now we've kind of covered the very, very general ideas of what's new. Do we have anything we wanna talk about before we dive into the, the features?

Just a quick follow up, Sabina, real quick on, uh, on your points on the three x switching performance, will you get into that in more detail or can you point to like, how are you really measuring that? That would be great. Uh, follow up reading.

Oh, that's been actually, that's me, that's been measured on our end by essentially pushing packets through our own kernel. So assume the number up until a certain point was 20. We basically saw that go up to three times more and three times more was actually the conservative number that our product team has ended up giving us as a result.

So as you do kernel optimization, one of the things that happens is how much, how encumbered is the processing of your code as a result of, again, all the instructions you're giving it. So this is more of an efficiency, um, report rather than anything else. And it's very easy to test because it's essentially you put an old version of this next to the new one and that's also how they've been testing to see is it getting better by essentially fluking it.

So that's great and it's a great story. It does sound more like a kernel optimization story than actually, uh, um, uh, increase in switching performance, if I'm hearing you right. Well, it is a current optimization in, uh, for the switching part.

So coming back to what's new for the next few slides, what we wanna talk about is, again, what's in there, what are the key features that drive a modern, secure and applic application ready platform? And you're, you're seeing these pillars again that we just talked about a few minutes ago. There are things that we've done or features that we brought that help you do infrastructure modernization.

We've completely redesigned the day zero. How do you get VCF nine installed? Um, we've completely, we've set you up with a new VCF operations console.

VCF is now multi-tenant, something that hasn't been the case before. Next up, it's a lot easier to troubleshoot. We've introduced enhanced diagnostic routines.

We've introduced proper fleet management that helps you with the different deployment scenarios that you might have and helps you upgrade quickly and lost that cost chargeback and showback that you have as part of the platform. On the app modernization bit, you have the consumption experience, you have cloud networking with VPC consumption exactly like you have it in AWS or Azure. And then self-service catalog infrastructure as code and VKS Cluster management.

VKS is our VMware Kubernetes service. It's our Kubernetes, um, deployment on top of VCF and it helps you do KS cluster management on top of VCF from the security modernization perspective. Three key features here, security operations dashboard, look at a glance of what's going on in your infrastructure, the configuration, compliance and monitoring and lost the identity and certificate management.

Below this, we have some of the more core features that we've talked about before. We kind of highlighted five of them. There are many, many features.

The payload is about 300 plus features wide. Again, ma major features, not patches, not fixes. So we talk about things like native recovery with vsan to vsan data protection with deep snapshots.

We talk about confidential computing as part of the core. We talk about maximizing throughput with VCF data path enhancements, the three times that we just chatted about, uh, cost performance with memory tiering and v send Global dtu. Quick question on the previous slide.

Uh, you had a few, and I think they show up on this one. Uh, basically a few features that, you know, there's an asterisk like they're available with an RPQ. What's driving that and do you expect them to be just available to any and all customers at, at some point, like the global, do you do confidential computing?

So a lot of the features that are available with RPQ, um, are designed for like, they're, they're ready to go. They're absolutely customer safe. However, there are still certain things that our engineering teams are working on that we'd like to make sure that they have a smooth experience with.

Like for example, they might not apply for 99% of the cases. So we want the RPQ process essentially. You can get the feature, that's not an issue.

We just wanna make sure that we track that we communicate with you, we understand where you are in the usage of this feature and we guide you towards deploying this. So it's more of a, again, especially when it comes to data focused features or things like confidential computing, we wanna make sure that the customer has a good experience with the feature. What happens is a feature is released with an RPQ asterisk.

We go through this entire process of requesting it through our teams and then usually by the next version or something along those lines, it's being released as GA with the feedback from that, um, from the previous RPQ session. Gotcha. Let's talk a little bit about the feature.

So we talked about infrastructure modernization. What does this cover a foster deployment with streamline operations, things like data simplicity. Again, a lot of marketing terms, but I actually want to show you what's in the box.

If you're familiar with VMware Cloud Foundation from the days of five to one before, I wanted to kinda show a little bit of a timeline of where we were before and where we are now. And you'll notice that there's a little bit of a, not necessarily an equal sign, but a growth sign of what was before the Quick Start app for VCF deployment has now become this full fledge cloud foundation installer in nine, oh, what was before a mixed lifecycle management has now become this performance fleet management that allows you to do really a lot of different things as part of your upgrade. The Offline Depot functionality that allows you to, um, work with dark side environments that a lot of our more military type customers are asking for has also been folded into this fleet management where you have a lot of options and a lot of ways to do upgrades and download packages and so on.

Then we had, uh, vsan Max and vs storage clusters as different offerings, offerings of vsan. Now, the way you wanna look at vsan is a technology in VCF that powers the way you store data. You want something like very, um, dense storage, fantastic.

Use that part of, of VCF and that part of vsan as part of VCF. You want to do basic primary storage with all the functionality attached to it. Fantastic.

That's, again, available to you all sorts of single sign-on. Like we've, we've introduced enter ID in the past. You're gonna see a plethora of options for a single sign-on nowadays and in general for identity management.

And then we've also done all sorts of migration tools in the past, like the easy adoption of virtual networking that basically allowed you to stretch an existing network from a legacy environment, from a layer two VLAN based environment into VCF and help you move the gateway over. This basically now becomes, there's now functionality to that at layer three level on top of the fact that you can turn some of these into VPCs. All of this is basically a growth story that by VVC nine you have all these featured, they're more matured and you're able to use them with automation operation tied to them.

Talked a little bit about the installer. In the past, it used to be that we had something called a cloud builder. You needed a JS O file, you'd load, there was basically an entire epic journey to go from what are my requirements and where do I want to go and how can I deploy my, my, um, my VCF in my, in my private cloud.

Now this has basically become a, okay, launch this, download this as a VM and start playing with it and start ask, answering some of the prerequisites. But by the way, notice how from the beginning it asks you, do I do I want to do an offline depot? What are my depot options?

If I already have A-J-S-O-N, I can go ahead with that. Do I wanna do a guided workflow instead? So it's up to you how you want to get started.

You have any number of options from the beginning and you have a very mature GUI to help you go through this. Or again, the more coded version via A-J-S-O-N script, um, once landed in this, you can start by deploying either VCF or VVF. Obviously here we're gonna cover VCF.

And if this is the first time you're running this at all, you, again, you still don't have those two options of I have an existing vCenter appliance and I'd like this to become a VCF deployment or do completely v uh, Greenfield and just deploy, uh, VCF from scratch. I have both options. Sign in and sign out.

Sorry, opt in or opt out from certain parameters. Now a quick question on the installer if I can ask one. Yeah.

Um, so with the, with with the ability to deploy to existing infrastructure, uh, I'm assuming there's checks to make sure that you're at a version that's supported of vSphere and vCenter and, and whatever else, what happens after that import? 0, um, is it just automated to upgrade everything that you're importing or can you stage it, you know, around whatever maintenance windows? Like that's the part that I haven't really understood when I've seen this installer so far.

So there, that's a two part question. So there are prerequisites by default that once, if, especially if you pointed to an old, older or previous deployment of whatever you might have, it basically goes and it looks something like this, right? So it checks, it has any number of checklists.

By the way, this has, this is a slightly older view of it. It actually is a little bit even longer nowadays. And, sorry, that's the one I was looking for.

Um, all of these pre-checks are being done to see, okay, does this work? But even before that, in order to consider an upgrade to VCF nine, you have to be on the correct version for your vSphere, your vsan, your VCF deployment, assuming maybe you're coming from A VCF deployment. 2.

As you're picking up NS X into let's say a VCF environment. So you have to put in the work of basically upgrading some of these and getting to the target version. Then what happens is an import and convert process of getting from the existing version or existing, sorry, bundle of products that you might have available and converting that into A VCF deployment.

Assume you start with the correct versioning. What happens next is you tell it to convert vCenter into a new VCF instance, you can point it to your NSX and VSENSE setup, and then it takes them basically all three, four, whatever. If you have old ARIA deployments, that's also something we can convert and then bring it into a VCF deployment.

And, and, and the reason I ask is like all of those components and upgrades, that could be a pretty lengthy and disruptive process. And so what is the recommendation for customers who might have a rather large environment with all those management components and maybe dozens or even hundreds of ESXI hosts that also are on, on a, you know, three nine oh version and they validated that it's on eight oh U three, but is it going to try to upgrade every single management component that you just outlined and every single ESXI host in sequence and you just have to wait for it to complete. I'm just kind of curious there, what that looks like.

So you started with the vCenter instance upgrade itself, and I think this might be a better conversation for, uh, one of the future sessions, otherwise we can really spend 30 minutes on it. But the, um, what we advise is inevitably do it in a maintenance window. The vCenter part does not, as you well know, does not create any disruption to the actual vm.

However, the, the bigger the environment, the more risk the same for NSX. If you bring in N-S-X-N-S-X manager does not necessarily disrupt traffic. Traffic continues to be forwarded, but do it in a maintenance window.

So far we haven't seen a lot of issues at this stage as we imported different V centers in, but I don't think I ever recommended to do it to in full production either. I see. Thank you for that clarification.

All right. From an existing infrastructure perspective, this is kind of how, what you were asking, like how does it work? You point it to your vCenter server, you point it to your NSX instance.

If it's not connected to one, that's fine as well. You have two options here. You can bring one that you've already had or you can deploy a new one.

And then as a result of kind of all the data that you feed it in, you go through this process of, okay, what are my pre-checks? How does this work? Are there any unsupported?

Um, uh, sorry, topologies. So for example, when we first launched this import and convert features, we weren't supporting certain type of apologies. Those have since been resolved.

2, but this would be the point where, okay, this is the configuration that we don't know how to transfer into VCF nine. We're gonna stop here and tell you to go either fix it, change it, or abort this migration. What happens afterwards is you took the different components, you took the different technologies.

Do you want the VCF operation instance? If this is a greenfield deployment, you probably won't have a VCF ops deployment yet. So what you wanna do is deploy a new one and attach from there.

But if this is your second, your third, your fifth import, then you basically say, I've already deployed my VCF operations. Please just attach to that. So Ken, to your question from before, if you have multiple deployments, if you have multiple V centers, what we've worked with customers to do in the past is basically take them one by one and slowly attach them to your VCF uh, environment over time and start consolidating them.

It, you don't have to do them all at the same time. You can do it usually cluster by cluster is probably the best way to do it, but again, it's a customer decision at the end of the day. Okay, cool.

For the VCF automation part, same story as for VCF operations. You start with, you have none, you wanna deploy one. There's an interesting discussion here where if you have existing area ARIA automation deployments, what you wanna do is do a migration post deployment at VCFA.

So if you're running ARIA Automation today in production and you have existing workflows, that's gonna be a separate migration and that's not going to be part of this ingestion process. You go through all of these steps, you've imported your existing infrastructure or deployed new, you've deployed VCFA, you've deployed VCFO, you go through all of this on a standard greenfield deployment, we're looking at about four hours for the full data center deployment. What happens next?

Your VCF operations is available to you. And that's exactly where we go next. It's just a question about the migration.

Yep. Say that automation, you don't get, uh, migrated existing IO automations. What about I operation For our operations?

It's, again, it's also a migration process where we can pick up some of the stuff that you have there, but you do it separately out of VCF Ops. Okay, got it. Speaking of, now that you've mentioned it, the VCF operations console is your landing spot, the mo.

So the moment you've, um, you've deployed VCF nine, this is basically your main console. It's not vCenter, it's not vSphere, it's not vsan. The whole idea is to walk away from managing these individual technologies and kind of leverage this central console for all of this.

One of the first bits that you can do is, okay, again, you have a new deployment. Let's start thinking about how we wanna do our identity provider. And remember, I, I told you that in VCF five two, we introduced support for intra id.

Now we see all of this, both modern and legacy identity providers that you can set up. So go ahead and set up how you want your users to log in and out. What else can you do straight at a glance or straight away, uh, out of your VCF operations, deployment certificate management.

And kind of to the point of what I can ask before, like how do I, how do I import different pieces of my current deployment? My many, many V centers, you'll notice here that you have different data centers and all of these have also different like, um, environments that we brought in. 0 1, 0 2, 0 3, all of these are different, different V centers that we brought in.

All of these you can bring in one at a time in the end. So all of this are different pieces of infrastructure that we've imported as a result of it. That being notice also the number of vCenters available here.

Coming back to the certificate management, this is also how we aggregate data across the different environments. Up until this point, there wasn't one central licensing, um, view for everything like NSX, vCenter vsan. So if you needed to, to see whether your certificates were aligned or, um, apologies, we're gonna need to edit this one out.

Um, centralized certificate management wasn't available before. And if you wanted to see your, your certificate rotation, your expiry dates and so on, you had to consistently log into the different platforms and kind of check when any of these are gonna expire. What we brought in is the certificate status from the different platforms.

We bring it in, we present it to you here, and you get notifications. You get all our renewal status, and you get to see like at the cross of different components, where do I have the most challenges at the moment? Where do I need to go and start thinking, okay, I need to start scheduling planning for this certificate.

Uh, lack of certificate rotation has caused a lot of issues in a lot of environments. So having this at a glance allows for, uh, IT practitioners to just go in and go like, okay, I need to worry about this. I starting with next week or next month.

One last bit. So we've, uh, we've had an entire beta program running for about seven or eight months ahead of our launch. And this was run by customers large and small.

And one of the things that we've wanted them to test was the installer and VCF operations kind of process feedback experience. And for some of these customers, VCF was entirely new. For some of them, they were existing VCF customers.

The feedback has been overwhelmingly positive, especially coming from kind of disparate environments and so on. The fact that it's intuitive, the fact that it's, um, easy to understand kind of what should I do next? How should I import this, where does it fail?

Has definitely been pieces of feedback that we've gotten that's meant a lot better than in the past. Real quick on that, Sabina, can you put any numbers against like time saved? Like what is the new installer?

You, you said about four hours for a full data center deployment. 0. So there's, there's a two parts.

So strictly how long, so the process that you see here, kind of all of this, when it gets from nothing to 100%, used to be also four hours in the past. Largely a lot of this is how long it takes to deploy all of the machines there. There's maybe some efficiency, some efficiency to be gained going forward, but it's largely about setting everything up correctly.

Where it has helped quite a bit is more at this stage. So when you, when you got started in the past, the cloud builder or the Excel that we asked our customers to fill in, in order to kind of get, get started, all of that was, let's say, from a product of a different era. And definitely that used to slow down quite a bit.

The fact that it's intuitive, um, allows for kind of the IT admin and the architect to go through this pretty quickly and start playing with their, their vCenter fleets. What I can share with you from customer, uh, discussions that we had in the past is customers are less concerned about how do they, not necessarily how do they convert and how do they deploy, but more like look at it, looking at their vCenter sea of, um, deployments and thinking about how do I wanna get started? So this has become a non-issue.

The four hour deployment time is something that you let it do on its own and so on. Where the efficiency becomes is creating that normalized architecture in a discussion with a large bank. What happened was they were a result of a large amount of acquisitions.

So they had a lot of different configurations for their V centers, right? So they acquired this smaller bank and it was configured like this with this kind of a security profile and so on. So when we were discussing migration scenarios, what became very obvious is this is an opportunity to deploy from scratch.

This is an opportunity to normalize the confi, secure it, make it compliant. Again, big bank problems. And, and that's what's actually been, this is an enabler of, because since here you can hook in whatever vCenter you want and bring it in.

Again, you have two options. You can bring it in with all the garbage, or you can start deploying new and kind of move your VMs onto the new platform. So from an efficiency perspective, I can't say that the installer goes all that much faster.

All of the steps from going from traditional legacy infrastructure to a platform based infrastructure, however, is an enabler. And it takes less time, Helps, helps ensure correctness. That's like one theme that I'm hearing.

And and just so you know, we're trying to get away from spreadsheets and, uh, for source of truth and network automation too. So, Well, you're gonna like some of the features that are on the roadmap, which my, uh, networking, my, my networking team would definitely complain if I start talking about it. But you'd love some of the stuff that's coming.

Um, I think it, a lot of IT infrastructure runs on Excel, especially when it comes to accounting for everything that's going on. And this is where kind of that centralized view is supposed to help. Hope that answered your question.

I know I went a little bit roundabout. All right, so we've gone through the identity provider. We talked about certificates, we've talked about the overall installer feedback.

This is the don't trust me, we have happy customer slide. And by the way, it's uh, VMware explorer. You're gonna see a lot more of that 'cause we're focusing on bringing as many customers that have interacted with our environment as we can.

Now let's jump on application modernization. What does this mean? It means we want to help you deploy, deliver, and automate.

So when you have to, uh, deploy a new application, there's any number of components and IT practitioners, it, um, architects don't want to necessarily deal with trying to understand everything that a developer needs, but giving them the building blocks straight out of the box becomes a lot easier for them to kind of piece it together. Remember, the primary reason why public cloud became increasingly popular is because somewhere a frustrated developer was able to request infrastructure services a lot faster than asking their IT team and 15 different tickets. So let's do that in the private cloud as well.

Similar journey map here and five two, we have, we had infrastructure as code services. We had, uh, Kubernetes support, we had container support. And what happened with that?

0, we've we're giving you that one consumption experience. We're giving you the VPC consumption that again, looks exactly like the public cloud. That multi-tenant, again, so many customers for so long have asked us bring multi-tenancy into the platform.

And on top of that, you also have tenant-based private charge back and show back. So if you wanna set up, set yourself up as a service provider for the rest of your organization, that's the way to do it. You can look at the cost of a certain tenant, you can look at the cost of an application based on the amount of resources it's using, and then you end up with things like Kubernetes as a service.

The automation console is the the one stop shop for, again, whether you want to do service deployments or you wanna set up things like, I want a new VM, or I want a load balancer, or I want a new VPC. Or if you wanna go even further than that and deploy applications from a catalog. All of these are out of the box.

This is at a glance how this looks like in VC F nine and we've had customer going like, I've seen this kind of thing before. How is it different than how it was before in VCF five, for example, or in our automation. One of the interesting things in the background of this, so even if this screen seems a little bit familiar, one of the things that makes, uh, the experience different is in the past, if you wanted to create automation between, let's say for deploying a vm, you needed to tie it to what kind of image do you want, how much storage and so on.

And all of those you'd have to do manually. There wasn't a hook that you can say, okay, go pull storage, go pull this, go load this image from the content library. All of this you had to set up manually for one vm.

Now it looks a little bit more like this, where you can basically point it to the resources that you need. What kind of VM class do you want? How many nodes?

What kind of, um, C-P-U-G-P-U configuration do you want? You can point it to your content library and say, okay, go load this image. And again, the entire automation flow is set up for you.

So when you publish here and you say VM service, all of this is curated by you, the IT admin, the IT architect, and you get to just choose between these and notice the versioning here. You can upgrade these over time. Like you can give them more or less options.

You can limit access, for example, to an image of red hat that you don't want your, um, your team to use anymore. Things like that. So the major differences speaking, Scott, you were asking me before like what's, what's the delta time reduction in between the fact that you don't have to do the work anymore?

In this case, I know your question was about the installer, but every time I kind of have these presentations, the question is more like, okay, what's actually changed? What's changed is the, you can set up a VM service in a matter of minutes nowadays, same for container service as well. And on top of that, if you wanna get a sense of how much does it cost, you can also put a price estimate on top of it.

You can configure your, um, your usage, how much does it cost, for example, for your memory, your storage and so on. And it presents to you with a total cost of ownership for that app, that service, that vm, that container VPCs are literally a one-to-one experience compared to the pub public cloud. How can you get your slice of cloud, a piece of networking that you can use straight away?

We've heard customers also wanting to use this as very, very soft tendency. Like you isolate certain applications or certain group of applications, you know, this piece of my biotech team wants to deploy this application, have A VPC. Some of our, uh, advanced services even allow you to do specific firewalling and threat protection within that VPC as well, or at least at the edge.

So you can prevent some of this. So you have the closest, you have security closest to the VPC ingress and egress. Again, nothing too fancy about this.

You get a bunch of information about how is it working, how is it used, and so on And on top of it, you can also do that cost management at the per VPC level, the big thing, the multi-tenancy, the benefits, I'm not even gonna start talking about them because we all know them. You can slice up your entire, uh, data center or private cloud deployment as you see fit across the different organizations. But what's even better is that you can see how it's being used.

Again, we're pulling data from the different parts of the platform. This is a little bit what the result looks like. And as I mentioned it about a hundred times ahead of this slide, the uh, the chargeback showback and cost control looks a little bit like this.

What's interesting with the cost optimization opportunity is that you can go further than just the tech that's in, uh, the VCF platform. So you can do modeling with things like, uh, the energy that it costs you to consume for that data, for that specific VM at that stage, or data center cooling. All of these are fields that you have to manually input, but it allows you to do a full cost modeling of, okay, I deploy this.

How much does it cost me? So it's dynamic when it comes to, you know, how much CPU memory storage am I using? It's static for things like how much cooling because that's an external source of information, but divided by the correct amount.

And then you get at least, at the very least, a general average on a perm per tenant per VPC level. And then you can publish this towards, um, either whoever the BU that's using this is or the general consumer. Speaking of, let's take a look how this actually looks like at the more talent level.

You'll notice how much am I using from a resource perspective, but also how much will it cost me? And the general trend analysis, what IT manager found interesting about this is that it allows them to do, um, their annual planning and their annual budgeting for next year based on, okay, what have we used and what's our capacity today? Am I just running a bunch of oversized VMs or can I go ahead and, you know, increase by 20% the same way I do every time from a security modernization perspective.

This is our last chapter, just a little bit more to go through. We've done a few things as usual in five two, we've done like in-service ESXI patching, we've done a synchronous upgrades and other data protection features. What happened in nine oh is we've introduced a security operation dashboard.

It's very much powered by VCF operations and this will grow more and more with the future releases to bring more security insights into it. We've done configuration compliance and monitoring. Things like configure drift come to mind, data and ransomware protection via the vsan, um, feature set.

But also, uh, through some of our advanced services application. Troubleshooting and insights fits into security modernization largely because again, if something goes wrong, we wanna make sure that it's quickly resolved and nobody else can take advantage of it. The security operations looks a little bit like this at the glance.

It offers you kind of what CVS are open, do I need to tackle this very quickly? And again, they're, we're tracking CVS across the platform. The host encryption mode is the next item.

User permissions, authentication. Have there been any failures? This is fantastic for somebody like a security architect or even a CISO to kind of look, is there something going wrong at the general glance and what should I be doing about this?

But the really interesting one is this config drift feature where again, you have a standard, especially for compliance reasons, you have a standard of how you want these to look like. Have we drifted, have we had somebody that accidentally introduced a new configuration somewhere else that either should have been replicated somewhere or it should have never been there in the first place? This allows you to send alerting emails, create tickets as far as, Hey, something's happened.

Are we aligned with our framework still? Are we still compliant? Can we go ahead or is this something we should go ahead and correct?

One of the last things I want to talk about is the application insights and troubleshooting where we wanna do better with troubleshooting in general. One of the questions that we've gotten in the past was, you know, I'm used to vSphere. I'm not a network engineer, I'm not a storage engineer.

Troubleshooting isn't the easiest thing to do for me when you give me this massive platform. Well, to kind of help with that, imagine something goes wrong in your whatever application, here's a wellness application, but it doesn't quite matter overall. We know what all the components are, we know how they're connected, we know how to reach them.

So what if we help you kind of troubleshoot, say there is a drop in latency today, or sorry, an increase in latency this morning. And we try to understand what happened here and why. Normally what you do in a more legacy environment is you start going and investigating.

Somebody logs in on a switch, somebody looks on the VM of everybody plays a game of hot potato, not my fault, probably networking. And then what you open a ticket management is still wondering what's going on. This allows you to do kind of full application mapping of what's going on in there and see, okay, it looks like we can't reach the loan balancer, the the web service, the whatever.

So even if it doesn't give you the full output of what happened, it gives you the very clear, these I can't reach as of this time. Like I'm gonna narrow for you kind of the time window when this happened. The first instance of me not being able to reach whatever it was is that time.

What do we do next? And then you can isolate and go from there and you can continue your troubleshooting without that game of hot potato by the end of the day. It shows us, okay, we have this much packet drop and these are some of the issues that we've noticed.

Again, simply allowing for, uh, the IT practitioner to play around and start a lot faster troubleshooting in the right place. With that in mind, I think this is a good overview and a good, uh, place to get started, kind of understanding VCF nine. I'm curious, are there any questions or anything you'd like to cover on top of this?

Uh, I guess my, my only question is for the customers who are not currently running VCF and they're planning to deploy it with nine oh uh, and they maybe feel a bit overwhelmed by everything you've shown today, 'cause it's a lot of new stuff for them, what's the best place to start? Should they just poking around operations since that's now the main point of administration for everything you've showed or certain features they should check out in particular, what do you think is this way for people to start getting value out of all this new stuff they bought? You know, it, it's great that you asked that because I think it's a, it's a common ask from a lot of our customers.

Like it's, it's definitely a massive platform first. Um, there's a lot of free certificate certifications and reading and so on. We've put a lot of materials out there.

So it start by deploying this in the lab, play with it. There's a lot of good information about how you can put together kind of your own deploy, especially if you're an existing VMware customer. Talk to your rep about how you can get a lab license and start playing with it.

And from a aware perspective, VCFA and VCFO are definitely the, the right place to get started. 'cause that's the part that's gonna feel overwhelming. That's the part that's gonna expose you to your entire infrastructure.

And then start replicating things that you do day by day. Like say your day-to-day job is just managing a massive VM environment. How would that look in VCFO?

How am I managing my vm? And then once you get comfortable with things, like what I like to do is breaking things, right? I start breaking it.

Does it give me the right alert? Does it show me where it's broken? And then play with it some more.

And once you get kind of comfortable with VCF operations jumps towards VCFA and start seeing like, can I build some of my own automation? I have this thing with Ansible in my infrastructure. Can I easily replicate it in VCFA?

I have some terraform deployment here. Can I connect the Terraform provider into my environment, my lab environment, and play a little bit with it and see how that looks like. And basically start doing the things that feel already familiar to you from vSphere, from vCenter, from vsan, from NSX, do them in V-C-F-O-V-C-F-A, start getting used to it.

And then for everything you have a question for, there's a ton of documentation about, Quick question about, uh, application, um, modernization perspective. So how really can reduce the complexity from this? Yeah.

Because yeah, from a platform perspective I, um, uh, some companies are thinking about the application, the deployment, and the, all the lifecycle, like, you know, the CICD pipeline and stuff like that. Mm-hmm. How this can be easily integrated in, uh, into, into VCF.

And I know, Oh, you've, you've brought a question that I really love. So, the nice part with VCFA is that it actually integrates with GitHub. We, we leverage something called Argo cd, and you are able to link, so you, you use GitHub as, as your source of truth and use Argo CD as essentially your broker for that.

So you're able to pull your automation into your CI ICD pipeline. So if, if you have that kind of customer, we have that level of maturity. The fun part is you can actually teach your developer teams to request infrastructure through your CI I CT pipeline.

Yeah. So say, say you want a VM and you want two petabytes of storage for some obnoxious reason, and you can only allow for something like, let's say one petabyte or a hundred terabytes and things like that. The nice part is that it's not only allowing you to go and request it through that pipeline again.

It'll make a call from GitHub to Argo CD to VCFA. What happens next is, what am I allowed to do here? You know, requester X has requested two petabytes of data.

I'm allowed to give it about a hundred terabytes. It provisions the a hundred of terabytes. It sends a notification saying, here's your a hundred terabytes.

Great. And as a result, you also get you the IT manage, sorry, the, uh, IT architect or vSphere admin. 9 that's missing.

What do you wanna do next? So you both have that level of control that if you put your, uh, blueprint in place, um, this is all we're gonna give you. But you also know like, Hey, there's a need for more.

What do you wanna do about it? And it's all part of the CICD pipeline from the developer perspective. They don't have to open a ticket, they don't have to talk to you.

They don't have to do anything. That's great. And the other thing is that is about, uh, the integration with, uh, the brownfield scenario.

So, uh, I was thinking about the integration with the enterprise, uh, software. Like, you know, you talk about the security improvement, security and also the, um, monitoring and the operation. So sometimes you, yeah, in a lot of company, uh, things are being tracked by, for example, by Jira, by Clash, and by the other software.

So is there any integration, native integration, or people should do this manually or with a third party tools or, you know, Best answer I can give you here is that it depends depending on the provider, depending on which version as in we're looking at integrating with more and more providers. I think it's really specific to what the customer has. And with some we do.

Yeah. And with some we don't. And with some we will in the future.

The idea is to have a open ecosystem where we work with our partners and make it easy to do, again, automation left to right, not just within our platform. So if it's not doing, it's, sorry, if it's not working now, it might be on the roadmap.