Freedom of Choice for Cloud workloads with Dell and HYCU

Uh, my name is Shiva. I work for the product management team at Haiku. Been with the company for almost six and a half years now, but my passion and my career is basically based on data protection.

Uh, I started my data protection days back in er, I don't know if you remember that. It's been almost 20 years now. So, yeah, I feel old.

But, uh, started my, uh, Deb, you know, as a support engineer. Here I am working for the next gen data protection company. As everybody, everybody pointed out, uh, now today I'm focused towards highlighting how Haiku can basically help with cloud mobility.

So to speak of, and when I speak of speak of cloud mobility, one thing that I want to get very clear about are the set of use cases. In this specific case, I'm focused on application-based workloads, hosted on VMs, right? Could be on any hypervisor or hyperscaler.

How our customers can basically achieve workload mobility across on-prem, across platform, and across public cloud infrastructures. Now, before I get into all those aspects, I just wanted to talk about the real challenges that are involved, especially when it comes to moving your workloads from one platform to another hyperscaler. And frankly, there are a lot of challenges and I've done my best to abstract them as much as I can.

But the first part that we noticed amongst our customers is when you have different cloud vendors, you basically have different set of rules and requirements, right? You have a different set of API calls and cl I to ensure how you go about converting your V disks. How you, how do you go about converting your VM types?

How do you go about specifying the network settings? How do you go about specifying the driver injection at the source or at the destination? How do you go about specifying the location?

How do you go about specifying permissions? The list keeps going on and on. I know this because I've been dealing with customers day in and day out.

But you cannot expect an IT ops or a cloud ops guide to remember, remember all of these steps, right? So that's the first challenge ever. And furthermore, to exacerbate this problem, you basically have vendors like cloud vendors literally changing their API calls.

Some of them are announced and some of them literally happen at the backend. Nobody knows. And we literally have to discover them by ourselves, right?

So that's the first one. Literally moving go goalposts. The second part is rising cloud costs.

Now, let's not fool ourselves guys. I mean, cloud is probably the most optimal solution out there, but they're not cheap, so to speak of. And when you take any business, I mean the cloud consumption or the rate of cloud consumption is always gonna go up.

Why is that? Because their business demands them to do so, right? Fair enough.

You can, you can optimize it, you can do whatever you want. But the cloud consumption is, the consumption is definitely gonna go up. And then you try to back them up with no optimization in place, your costs are going to go double, triple, or even quadruple.

And I've seen this personally amongst our customers because we thought object storage was the cheapest. Think again, it's not right. And last, but not the least, cloud in itself is not immune to inflation.

Pun intended, of course, right? Because I don't know how many of you remember Google Cloud announcing that they were going to increase the price of snapshots. They were not kidding guys.

They were true about it. And that's when Google Cloud really start to look at solutions like Haiku very seriously, because we are one of the few solutions who support a broad variety of Google Cloud workloads with no snapshot dependency, so to speak of, right? So that's the number two reason rising cloud costs.

Number three is about combating cyber threats. Now, I think this is more of a dead horse that we've beaten over and over again. The bottom line is nobody's immune to this.

Nobody's unimportant, right? Everybody is basically vulnerable to cyber threats. And you have several forms of cyber threats.

Number one is accidental deletion. Now this is something that is kind of laughable and nervous at the same time. How many of you remember Google Cloud's incident with uh, UniSuper based out of Australia?

So long story. Yeah. Okay, I see some nodding, right?

We got a native here. Yeah, what happened, right? Someone accidentally deletes a subscription and what happens pretty much the entirety of the infrastructure gets blown away, including their backups.

Why? Because the backups are within that subscription itself. So it's accidental, but the damage is still the same.

Number two is insider attacks. We've heard several cases of disgruntled employees going rogue, deleting mailboxes, deleting data just to prove their point. So this is something that is unavoidable at the end of the day.

And last but not the least, my favorite malware and ransomware attacks, of course, right? So this is the third aspect, especially when people talk about moving their resources across public cloud because they want to maintain that security, right? So with that being said, how does Haiku and Dell together can help you out?

And again, I've tried to do my best in simplifying the story. First part is about innovation, right? Haiku has, as David pointed out, has really good innovation when it comes to discovering multiple data sources at an extremely short turnaround time.

Let me just get into the details of what that means. What I mean by that is, I'm sorry, the clicker is kind of weird. I should have known this better.

But anyways, alright. So what I mean by constant innovation is Haiku does have the ability to integrate with multiple data sources with high-end API level integration. What this means is we are enforcing all of the best practices.

We are ensuring that there are no performance degradation, so to speak of. And the connection in itself is always secure. That's number one.

Number two is we have the ability to streamline all of those manual steps that I just talked about when it came to cloud migration, cross cloud or cross hyper MI hypervisor migration. So you as an end user don't have to worry about keeping of those steps like manually. Number three is also keeping a constant tab of what changes are being made.

This is by, by default the most difficult thing to do because keeping track of changing CLI and changing APIs when they're not announced is the most frustrating part. And we do that real hard work. Last but not the least, our solution is extremely simple and easy to deploy.

You don't need any plugins, you don't need any agents. Deployed instantaneously requires minimal maintenance and absolutely requires no training so to speak of, right? So that is how we start.

We basically solve the problem of ensuring that we don't have any issues in integrating with multiple hyperscalers and hypervisors, right? The second part is DDV has proven optimization. Now, again, as David, uh, now I pointed out data domain has been in the field for a while.

I know this personally. I'm an ex NetApp myself. So, um, data domain is very close to my heart too because data domain is the king of deduplication and there is no other vendor in the market in the world today who has the ability to beat that number, right?

Point blank. But my favorite aspect about data domain is DDVE. That is the real crown gel people because DDVE is basically the virtualized version of your physical data domain.

And virtualizing a physical appliance is no easy task, right? Rubrik has tried to do this, Cohesity has tried to do this. Even NetApp has tried to do this.

And the rate of success is not even remotely close to what data domain has done in the market today. DDV has been in the market for more than 14 years. This is other than just the stats, what David North showed about with regards to the amount of capacity or the number of customers, it's about how long have you been in the market.

And DDV has come a long way, which is why it's one of my favorites, especially when it comes to optimization efficiency technology because with DDVI now have the ability to ddu the data on the cloud and also have the ability to transfer minimal amount of data for ingress and egress, right? And last but not the least, one of the best parts about DDV is it could be easily deployed on any of your hyperscalers or hypervisors, right? I basically did one of that yesterday, uh, on Google Cloud myself.

And it's seriously, it's not that hard at all. It's readily available on Google or any of your hyperscalers marketplace and the deployment is pretty simple and straightforward. So with that being said, I know I've been talking a lot.

Any questions? Alright, so that's the second part with regards to optimization. Now the third part is with regards to the whole aspect of how Haiku and DDVE can come together in terms of data security to combat cyber threats.

Right now, haiku is well known to provide air gap backup security. I mean other vendors call it as zero, uh, zero trust architecture. And this is something that's standard across all of the backup vendors.

Multi-factor authentication, network segmentation, multi-tenancy with role-based access control treating solution as a black box. And list keeps going on and on. We're all there.

I'm not gonna basically say that we are unique in this. Pretty much every vendor does this, but what is really unique is with regards to haiku and DDVE making use of the DD boost protocol along with the inbuilt worm functionality that basically caters to compliance and governance modes. So what this means is when you have DDV deployed on cloud, the entire traffic is proprietary.

You're not moving the data using NFS or SIFs. You're moving the data using DD Boost. While the technology in itself is inherently efficient, it also caters to being highly security conscious because it's no one on earth who can breach that traffic because it's totally alien to the rest of the world, right?

And in addition to that, you also have worm capabilities. So even if someone basically manages to read, reach that backup, they won't be able to delete it at the end of the day, right? That's number two.

Number three, our solution does have and built anomaly detection capabilities, which basically help you to validate your backups. And we can also help you to prevent from good backups from being expired by policy definitions. And last but not the least, haiku and DDV can also provide you with plethora of recovery options.

They could be granular, they could be full, or they could be a combination of both. And you can restore wherever you want. Alright, so what am I really trying to say?

With all these things, we not only help you to protect the data, regardless of where it resides, haiku and DDV can ensure that your data is truly versatile. And this can basically help you to prevent cloud vendor lock-in how? Two simple reasons.

We are truly hypervisor and cloud agnostic, meaning that you have the ability to back up and recover every anywhere. But in addition to that, DDVE helps us to achieve that level of cloud efficiency with regards to storage and egress costs. So you get to swallow that poison bill, but it's taken in a very gradual way that's not going to kill your wallet, right?

And that is how we basically cater to the cloud mobility story. Any questions? Just one?

Uh, sure. I'll say a note that I know we've, this is still a very real thing, but it's becoming less of a thing. 'cause we, we've latched onto the cost of egress fees.

Two of the, you know, top four have eliminated egress fees. So it becomes a bit of a fuzzier message as we go forward. I know there's more to it than that, but it's, I just, I worry sometimes a lot of, we still get stuck on that thing and then a cloud admin goes, no, I don't, there are no egress fees anymore, right?

So they there are hidden Behind it. It's just one thing to note that that message Is egress is just one part. I totally agree with you, right?

Yeah, yeah. That EG is just one part of the puzzle. Storage costs is basically the big problem out there.

Yeah, yeah. Especially when people say object storage is cheap. Uh, no, it isn't.

Even like if you go to the lowest tier, no it isn't. We've burned our fingers with that and which is exactly why we're partnered with a solution like DDV. But yes, I totally agree with your point there.

Uh, in addition to egress charges, yes there are a lot of hidden costs as well. And by making use of efficient data transfer, you also have the flexibility of reducing your, uh, you know, expectations on SLAs. Your RPOs and your RTOs are going to be way more quicker.

So there are more advantages to that as well. But yes, totally understandable. Alright, so I'm just trying to figure out what de and could have been partners for a while.

Was this one of the announcements? Have you taken it up a notch just recently or is this something that's been in place at the level, current level for a while, David? Alright, so, so um, KU became A boost, Haiku became a Boost Ecosystem partner last year.

So basically we're able to take the boost API and be able to use it directly for the communication that Shiva was talking about. Um, we've been working with Haiku as well as a number of other backup vendors over the course of the years. As I said, our, our priority number one is to grow the data domain footprint.

Um, but what took it up a notch is that, you know, we recently announced that we are, um, going to add Haiku to the price list and the ETC program, which basically allows our sellers to go and say, Hey, we're never gonna lose a DD opportunity for lack of workload support. If we walk into a customer and they say, I need to back up Atlassian, we're gonna say, Hey, guess what? We got it.

Um, so we've always got support for every workload. And how long ago was that? This was just beginning of this year.

Okay. Thanks. Thank you David.

Alright. Uh, actually it's a perfect segue. Okay.

Alright, Here we go. So we basically have one of the, uh, customer use cases, and I wanted David to basically talk about this. Thank you.

We just, uh, build, build it out all the way. Oh, sure. There we go.

Thank you. So, I mean, you talked about egress and, and you know, the importance of keeping that, you know, does it matter so much about keeping that line of communication between the application and the storage at minimal? Um, in this particular example, this is a, uh, federal customer.

Um, I'm not gonna go into who they are for obvious reasons. Um, but this is a agency or a group that have things like disconnected sites, forward locations, uh, where you don't always have full connectivity back into a home base to basically be go, go and replicate. So the, the more that you can minimize how much is going across the wire, the better.

And especially for these kind of locations that don't have giant pipes, connecting them back to home. Now, as I said, this particular customer was in the process of one, uh, bec migrating from VMware to alternative hypervisors. And what Iku was able to help 'em do, I said, we're gonna provide the industry's most resilient and cost effective storage for your backup, uh, software, for your backup data.

But we are gonna partner with software vendors to make sure that customers can adopt any workload if that means that they're moving from one workload to another. And in this particular case, haiku could help them migrate from VMware to the alternative, which in this case is an Azure based solution, uh, a, a Microsoft based solution, then that simplifies their life and helps 'em do the migration. There's a huge, um, um, transformation going on inside of this particular agency.

And what's more is that, um, when we look at, you know, the security posture, this is a very security conscious customer. So they care about things like, do you have your federal certifications? Are you encrypting everything?

You know, basically how buttoned down is it? And so the combination of haiku's, federal security posture, as well as uh, the security posture of the data domain product itself, which is a PL certified, um, is what makes this a super interesting solution. And at the end of the day, this is a customer has a lot of data.

So having uniformity across forward locations, central locations and cloud-based locations, the kind of security that we just talked about, ease of migration, ability to onboard new application workloads really quickly. Um, it just made a lot of sense. And of course driving down cost.

So, you know, one thing that Shiva mentioned is cyber and well, it, it isn't listed here and we, we don't talk about it as much, but look, one of the nice things about the data, the main product is a storage product. Typically when you back or when you any storage product, you have your primary, you have a disaster recovery copy, and then now there's a need for a third copy. What's the third copy all about what Shiva just mentioned.

If you've got, if you lose your account, if your account gets compromised, the third copy. And what we can do with data domain, and it works with iku too, is have a cyber recovery revolt where that cyber recovery revolts petition another account with a different set of admin privileges, uh, so that the administrators of the primary and DR account cannot get into it. And then a third copy can be used for recovery in the case that you were compromised in your first and second recovery.

So this all critical stuff, but I thought this would be a great example of, of, you know, sort of the, the importance of having this kind of a joint solution. Alright, thank you Dave. Just on.

Sure. Where do you see the, as far as like the percentage of people that are KU is add on to data domain more than data domain is a great add-on to a Haiku client. Like I would see that hiku as a different opportunity to be sort of like mid-market move down market, whereas data, data domain clearly of fantastic enterprise play.

So do you see bi-directional opportunities for Dell moving stuff into, uh, a hiku, uh, you know, customer environment? Look, I, I, I, I first, first of all, let me, let me challenge the notion that it's only a up a high-end enterprise, uh, 25,000 customers. Mo mo the majority of that number is actually not in the large enterprise.

'cause there's aren't that many large enterprise customers. What DDVE allows us to do is to go down, down into very small deployments. Right now we sell a ton of data domain used to be 3,300 model.

We now have a 64 10 model coming out that starts at 12 terabytes and goes to like, call it a hundred to 150, a hundred, uh, 200 terabytes. A ton of those go out. And you think about regions like in Europe, uh, AsiaPac, you see a lot of those kind of small ones going out and they're highly channel driven.

Um, what the DDVE offer allows you to do is just go take a vanilla bare metal off the, you know, cots commodity off the shelf server drop DDV on it, that now becomes a data domain appliance, right? What what Haiku does is it basically gives us the workload access, whether it's at that small or, or, or actually we're seeing a lot of this in even larger customers and we'll give you example of some, some larger ones later on. But some of the initial opportunities that I'm watching roll through, they're, they're across the spectrum.

They're everything from really small to really big. Yeah. And when I've been really big, I'm talking like, you know, in the Fortune 500.

Yeah. And of course we, in this particular customer case we mentioned as your local, uh, how do you do, do you see at AWS outposts appearing much in some of your environments? I don't see AWS outposts as much.

Azure stack definitely comes up. Yeah. Yeah.

And sorry, I don't mean to ask you a generic market question Yeah. Here, but I just saw that and it reminded me that I haven't seen, if You're asking me personally, I don't see them as much. I don't know if the Haiku team sees 'em as much.

Yeah, yeah. I mean this, this specific story does have a lot of remote offices, so to speak of Yeah. Uh, not exactly AWS outpost, but yes, we do have a lot of remote office branch office use cases in our case as well.

Alright. So back again. Um, so what I wanted to do was not just bore you to depth with slides, but also do some live demonstration of what our product is capable of.

Uh, now this is not like a really elaborate demo, it's a very simple architecture that we have. Uh, what I'm trying to do here is I'm just tr trying to showcase how our solution basically integrates with data domain on-prem could be a physical or could be a virtual, really doesn't matter, but it's on-prem. Uh, having the ability to basically do app consistent or crash consistent backups directly onto data domain.

And once that is done, how does, how do we basically make use of data domain and Haiku to basically have those backups replicated to cloud and even to restore to cloud? And in this case, I'm basically making use of Google Cloud, so basically trying to showcase the whole cloud mobility aspect with haiku and data domain put together. Right.

Okay. So let me just do a quick switch. We start with the primary haiku controller that I have, which is on-prem.

So this is basically a Haiku virtual appliance. Uh, the entire solution is basically deployed instantaneously because we deliver our solution as a package. Uh, it could be a QCA image, if it's a Nutanix, it could be an OVF file, if it's a nut, uh, if it's a VMware, so, so on and so forth, depending on the hypervisor.

And the entire deployment in itself is so simple and straightforward. The where the only manual hard labor that you have to do is basically discover your sources manually. And in this case it would basic, I'm sorry, I have a lot of interference here.

Uh, this is not the window why I wanted to open. So the thing that I, uh, was talking about is, uh, the only hard work that you have to do here is basically discover your sources, so to speak, of in this case, I've basically discovered a Nutanix environment. And in addition to Nutanix, I also have data domain as a backup target, which basically helps me to showcase the fact that we not only have the ability to protect, uh, data domain using your traditional NFS or SMB, but my favorite, we have the ability to make use of DD boost protocol for efficient and secure data transfer from source to destination.

Right? Uh, so this is basically one of those examples. By default, haiku has a plethora of integration points.

With DD Boost, for instance, we can make use of DD boost interface groups for load, uh, load balancing. We have the ability to make use of encryption. We also have the ability to make use of warm functionality using governance or compliance modes.

Uh, and there are other features that are on the roadmap that are going to be upcoming very soon. So these are the capabilities. Now, when it comes to performing backups, haiku is purely a policy-based data protection solution.

So what's happening here is we are basically making copies of one of my VMs that I wanted to showcase, which is the VM oh oh four. Uh, this VM is basically being replicated or being copied, backed up through your data domain. And again, you have tons of options here.

You have the ability to perform granular file level restores under restore file section, or you basically have the ability to restore an entire virtual machine. And this is basically the place where we can showcase the fact that it's not just about restoring VM to its original location, but we also have the ability to clone this VM onto an alternate hypervisor. So basically having the ability to not just clone the VM, but also v disks onto an alternate hypervisor.

But here's where things get even more interesting because these backups are basically being replicated onto my Google Cloud infrastructure, where we actually have a DDV running natively on Google Cloud right now. I also have a KU controller running on Google Cloud. And, and this is basically the Haiku controller running on Google Cloud as a DR solution.

And from here I can actually showcase the backup of that same virtual machine that has been discovered, located on your DDV running on Google Cloud. And from here what I can do is I can simply choose that backup and restore it onto your public cloud vendor. And in this case it's going to be your Google Cloud.

As you can see, I also have other plethora of options where I can, I can do restores to AWS or Azure or Azure gov or AWS gov. But in this specific demo, I'm just focused on Google Cloud. So when you basically do restore to Google Cloud, you can actually see that I can specify all of the details that corresponds to my Google Cloud account.

So remember all the checklists that I was talking about, uh, understanding the location, understanding the project, uh, I'm also gonna showcase how the VM types and, you know, the network configuration basically comes into picture. So I'm basically trying to do a restore to a US central region. So once you do that, this is basically the place where, uh, the solution in itself will be talking to the APIs of your Google Cloud.

That helps you determine, uh, what are the available VM types for that specific, uh, option. What are the compute and the memory settings that would be translated onto that. Uh, for instance, what I'm gonna be doing here is I'm just going to specify a simple name, alright?

And take a look at this. These are the, basically the set of options that are natively available on Google Cloud. And this is what I was talking about, the ability to ingest all of the available options on Google Cloud and make them readily available through the API or through the user interface from here.

What happens as we also have discovered automatically the VPC that has been configured on your Google Cloud. And all I would have to do is specify the appropriate operating system and just do a spin up. So now this has literally started the operation of moving an on-prem VM onto your Google Cloud infrastructure.

Yes. So, uh, those options, it was just Windows and Linux. Mm-hmm.

But like, what if I'm running Cintas or buntu or Windows Server 2020? It basically comes under Linux operating system, uh, itself. Uh, we basically try to make a broader classification of what kind of operating system do we run on the public cloud vendor.

Mm-hmm. So Linux basically can kind of, I mean, covers all of the flavors, so to speak of, But, but it's a backup, right. Of a system that already existed.

Yes. Right. So What, it's basically a parameter that you have to specify when you're launching that VM on the specific hyperscaler, so to speak, of it's more of a representative point.

At the end of the day, the backup actually does store the appropriate operating system. Mm-hmm. But it's more of a representative parameter that you have to specify when you're spinning up a VM on.

But why Would I need to specify it? If you know what it is, I'm sorry. Why would I need to specify it?

If you know what it is, if it's a, it's a Windows host and you know, you backed up a Windows host, why would I need to select Windows there? Because It's a parameter that has to be manually specified on the destination, so to speak, of, because it's a virtual machine based operating, uh, virtual machine based backup. So at the end of the day, operating system at that level is a moot point.

S You wanted to say something or, okay, I was just gonna say, I mean, if you know that it's Windows, why wouldn't you just specify that for me rather than me having to Yeah. Select it. I was wondering the same selected in the ui.

Why wouldn't you just automatically tell the hyper tell the cloud provider that? Yeah, it's Windows. Yeah.

So one of the reason I think, uh, people, we do actually ask for the operating system here is that if customers want some time, want specific disc alone, they wanna say bring the disc into a Windows machine. You're gonna, you can take a Linux disc and bring it up on Windows machine, things like that. That's why we give you an option.

That's why there are times where you can actually do cross operating. There is A weird situation where a customer asks for it. There Are think she actually brings specific, uh, data sets On, on the, on the, um, knowledge of the backed up system.

Question, um, the dropdown of machine sizes, do you recommend or suggest a specific size based on what you know about the backed up vm? Yes. Yes.

The, I backed up a two CPU 16 gig, And we select the appropriate type by default. Okay. Right.

So the thing, the reason why we give you additional options of VM types is you can either under or overprovision your VMs on the cloud, but yes, by default we basically select or provide you a selection of a default VM type based on the moment of compute and, uh, memory resources that, that has been assigned onto that virtual machine. Okay. How are you handling networking?

So networking, yes. That is one of the biggest items, right? So when it comes to networking, uh, we have the ability to discover all of your VPC settings that have been specified on the cloud today.

We basically request our end users to have those settings specified on the cloud itself within the given subscription. Mm-hmm. Uh, we as, uh, like we have a, a dedicated team of engineers and professional services engineers who can help customers recreate these VPCs for people who are brand new to cloud.

We have done this on Google, Azure and AWS. Mm-hmm. Uh, but for now, this has to be created manually on your cloud and it'll be automatically discovered on our KU solution.

I was thinking more the networking of the virtual machine itself. Okay. So are you able to, um, re IP that virtual scene as comes Up?

So we do, we do have the ability to do DHCP or even manual IP assignment if required. Mm-hmm. But yes, uh, we do, we can discover the virtual network interfaces that could be attached to vm and in addition to that we can do DHCP or a manual IP assignment if that is what is required.

Okay. Is there an orchestration component here? If I wanted to bring up a group of virtual teams?

Yes, We can. Some Time We can for like a DR test. So We basically have, unfortunately I don't have access to that user interface.

It's behind the VPN, but the thing is, we do have the ability to, uh, group multiple set of VMs based on any dis based on your discretion, right? Mm-hmm. You can create something called as a DR plan that specifies all of the rules on how the VMs are supposed to be failed over.

And this can also include all of your network settings, which would basically be a part of the larger orchestration operation on moving VMs from point A to point B. Hmm. Okay.

Yeah. Cool. So why you wanna say something?

Yeah. And in, in terms of, I guess expanding on that, in terms of the haiku controller, we've got one in each environment at the moment. Is there a centralized management interface or are they very sort of separate?

So yes, as far as centralized management interface is concerned, the answer is yes. We do have it for all of our public cloud and for SaaS offerings from a on-prem standpoint, um, we are working towards it. We are basically ensuring that we will be, uh, having more of a high level manager of managers view that would basically provide you access to public cloud and also to your on-prem infrastructure.

We are in the process of making that as a default option as we move forward because as much as a single UI sounds really good, it could present a weird set of operational challenges for people who are in a hybrid cloud situation. So we want to be extremely mindful about how we deploy that. Uh, but yes, today we do have that option for customers who need that, and we are basically in the, um, we have made some decisions and we'll be making that as a default option pretty much by the end of this year.

One thing that I'm seeing more and we're hearing more about, especially with VMware making, uh, changes that have triggered some different customer behaviors mm-hmm. OpenStack is becoming an attractive landing spot for a lot of enterprises, especially, you know, managed OpenStack and some proprietary tools. Do you see that that's a place where you could now in the same way we're doing to AWS and and Azure, is there capabilities or a plan to have a, a port to OpenStack capability?

Uh, AYA, can you answer that question for me please? Sure. I can add to what, uh, she was today.

If you look at it, we actually have one of the broadest coverage we actually support from a hypervisor perspective. We support VMware, we support Nutanix, A HV, we support, uh, Azure Stack, HCI and all the major public clouds we support. OpenStack is one of the things which is getting, definitely get a lot of attention in a lot of customers when we talk to a lot of people that a small number of production, a lot of people are kicking the tires, especially because of the VMware thing, announcement doing that, we are absolutely working with some customers and it's definitely in our roadmap.

You should see it literally in the spot. Yeah, it's interesting to explore as I know it's one of those things that it's, uh, and again, uh, I don't want to talk about the ex-girlfriends that, uh, is living with somebody else at the neighbor's house there. We've got Dale in the room.

I want to be respectful that I know it's tough to talk about that as a, as a customer, when I look at, you know, price pressure, totally, those things start to look attractive. And the first challenge is always how do I get my workloads there and how do I protect them once they're there? And I feel like that's OpenStax missing that in the biggest way.

And I see that your application awareness is helpfully would be amazing to see in that ecosystem. Look, if I, if I can add just the, the Dell perspective, VMware is, uh, is still and probably will be for some time the dominant virtualization hypervisor provider. Absolutely.

Yeah. But having said that, with with, you know, the announcements that came out with the way that the Broadcom acquisition went, um, there's no discussion that we talk to anyone, don't care large or small, where the conversation around alternative hypervisors doesn't come up. And, and even if they're not moving today, and even if they're not moving next year, because it can take time, every one of them is looking at the contracts that they have in the either one year or three year licensing agreements that they have is thinking about, do I have a dual vendor strategy?

Do I have ability to actually move if I have to? And so the conversation is really, I can't, I can't engage with someone who's not gonna provide me alternatives. And so I, I think Haiku providing all of the alternatives is great because they take away all that objection, basically, where do you want to go to, you have options.

Yeah. So from our perspective that that's, it's really important that we have that kind of coverage. Fantastic.

Thank you. Yeah. And I know it's a, it's a tough question to have to ask, you know?

Sure. Uh, so thank you. So a question on the conversion.

Yep. Um, are you doing a preconversion and then using the APIs to ingest it into the cloud? Or are you using the clouds, uh, vendors specific API conversion tool?

It's the latter. It's the cloud vendor specific APIs to basically make sure that we have the ability to not just move the VMs to the cloud, but also make those VMs portable. So we have to make use of the appropriate cloud vendors APIs for us to convert that VM from point A to point B.

Now, a follow up to that would be, um, have they gotten better over the, over time? Because I, I remember like AWS as an example would be kind of at will, um Yep. And not necessarily a high priority inside of AWS's environments.

True. So is that still the case? Have you been working with the hyperscalers to kind of get them to boost something for that particular ingestion process?

What you bring up is an excellent point because, uh, we try to use the native cloud vendors APIs as much as possible, but to your point, they don't always support the latest drivers because they come up with new, new, uh, CPUs and so on. So there are things there. We have to inject additional drivers.

So one of the thing our team does is tries to use the native as much as possible if it's supported, if it's not supported, we do actually inject the right set of drivers and do the conversion. So that is something our team has actually had to handle all the different varieties. Okay.

Absolutely. We have done that. And just, and one thing I'll say often we would say this in the offline after to say thank you, but I really appreciate that it wasn't now with a hundred percent more ai, you know, now ransomware as a service, what you've really shown, I like that it's just fundamentals plus fundamentals that literally are one plus one equals three because you augment each other's capabilities.

And I know within hiku, especially like the application level, discovery and intelligence is unmatched, you know, to, to what I've seen. So what do you see as the, the next thing inside hiku? Like what other workloads are you looking to stretch into?

So, that's a great question. It's probably, okay, so one that's a great question. One of the thing which, uh, we will actually on the next session, which is one of the thing we're gonna talk about is around ai, somebody you ai, but it's not about the AI from the context of saying how do we add ai, which we already, there are a lot of areas which we do.

This is around customers, as Simon mentioned earlier, some of the things that customers have, tons of data being everybody is working on AI project, right? They said 80% of the businesses supposed to come outta the AI solution before the end of the year. And that being the case, how do you keep that data safe for customers?

I think you heard also from David earlier. So that is some, that's the next topic and we will definitely talk cover a lot more about there. So, uh, just to do a quick refresh on the demo, we kind of got lost there.

Uh, the migration job, as you can see, was successful enough. Uh, uh, this is the job log and I was trying to locate the VM and it's right here, sorry, uh, bad eyes. So this is the Haiku VM uh, dash 0 0 2 that we just migrated all live, and the VM is basically up and running as we speak.

So that is the true power of our solution. The ability to basically move from one platform to another platform by literally streamlining all of those hundreds of tasks with no issues and showcasing this in live demo in itself speaks volumes. So that is the true power of Haiku and DDP put together.