Name: Beyond Legacy NAS: How to Secure, Simplify, and Scale Unstructured Data
Uploaded: 2026-11-11T18:01:18.432Z
Duration: 45 min 8 s
Description: Beyond Legacy NAS: How to Secure, Simplify, and Scale Unstructured Data

Transcript for "Beyond Legacy NAS: How to Secure, Simplify, and Scale Unstructured Data": Welcome, and thank you so much for joining our, joint webinar between Rubrik and VAST. We're so excited you're here today. We thank you for sharing the time with us. We hope you gained value from the conversation and the partnership between Rubrik and VAST. But we're really here today to talk about Beyond Legacy NAS and NAS infrastructure and how to secure, simplify, and scale unstructured data. It's a problem a lot of customers are seeing in the market with rising hardware costs, complex environments, and then the the surge of AI attacks on their data and their critical IP for an organization. We're here to hopefully give you some information on how a joint partnership between Rubik and VAST can help solve some of those problems. My name is Ricky Clopton. I'm based in Houston, Texas. Been at Rubik two and a half years. Before Rubik, I was selling file systems and then working in the partner community as well, but always had a passion for unstructured data and the the storage behind it. I've had the pleasure of running our go to market team for Rubrik around unstructured data for the last, six months. Before that, I was a seller, but really excited to talk to you today. I'm gonna kick it to my partner in crime, Jeff, to do an introduction as well. Hey. Thanks, Ricky. Yes. My name is Jeff Harbor. I have been at VAST a long time by VAST standards. I've been there almost five years now, And my function there is to to help in our partner community, really driving the strategic alliances of whom Rubrik is one of our most strategic partners in the data protection space. So my career has been all about data protection, in terms of backup recovery, Doctor, all those type of things. So this conversation actually blends well with kinda where I was and what I'm doing now in terms of with the the vast side of things, which is making sure that our customers understand what options are out there to protect workloads on VAST. And then, obviously, working with Rubrik here to help our customers understand how those two connections work well together. So if you paid any attention at all to the vast story, over the last couple of years, you've seen a shift. Right? So five years ago when I joined, we were a universal storage platform. That was kind of our marketing message out there in the in the world. And the good thing is when we found it ten years ago, the vision was this thinking machine concept, and it really was a forecasting or foreshadowing of what AI has been. So we've developed what we call the vast AIOS, which is designed to help support the AI workloads that are out there today. But when we think about it, fundamentally, we're an infrastructure component, and the things that make up that component live in the storage world and the compute functions for AI. So when we look at kind of what's out there today, a lot of the legacy platforms are challenged. So if you think about your infrastructure today, can it meet the needs for the next generation of growth and performance that's being driven by those AI workloads? Chances are no. A little bit as to why that is, probably a little bit as to why we believe VAST has solved that problem, why we're in nine out of the 10 model builders out there. Right? And vast majority of the NeoCloud providers out there. Because we built something that's kind of unique in the market, and it really has solved the question of, can we be ready? Absolutely. We're leading the forefront there. But that does create some other challenges for us. Ricky? Yeah. Thanks, Jeff. And then, if you're not familiar with Rubrik, I'll give a quick background of Rubrik, but, we're really the leader here in cyber resilience. We believe we've we've created that market, and and we strive every day to secure our customers for the next generation of data growth within their environment. A lot of the the the buzz today in talking to customers is getting ready for the next phase of AI and AI evolution. But to simplify that, specifically around unstructured data, unstructured data in an organization has been far looked upon as kind of the the the junk drawer of an organization's data state. What is it? Where is it? What's in there? Are there cat photos? Are there HR documents? Is there core IP? Is there engineering documentation? What is it? And we really partnered with VAST, and we go to market today, helping customers understand the data and the criticality that it is to their business. But answering this question, can you recover it? We're seeing a lot of attacks in the market that are focused on unstructured data and targeting the critical IP that your organization has behind it. And a lot of the times, your infrastructure, your technology, your security tools aren't built to ensure that you have a limited downtime in case of a recovery. And and Rubrik and NASCO Direct, the product that I support and we'll talk about today, is really built for you in your hundreds of terabytes, multi petabytes, millions and billions of files, and ensuring to the business that you can recover quickly and keep your minimal viable business up and running. So what you'll see today, in partnership with myself and Jeff, we're gonna talk about how you accelerate ransomware recovery. Ransomware has been a a big, big problem for a lot of organizations over the past few years, and now it's leading into, AI recovery and seeing those those AI threat attacks, but ensuring that you have an immutable air gapped protected copy. And then not only just ensuring that you have that third copy that's air gapped and immutable, but then what tools are you leveraging within threat hunting, threat protection, anomaly detection, sense of data classification to quickly identify clean restore points. That's the big, big if is not only can find what gets in and what they took or what the actual trade did, but then how do you recover cleanly? And then a big thing that customers are talking about is we have all this unstructured data, and all of our organizations asking to use some of these new AI tools. And how do we ensure what is clean and what is available for them to leverage AI against? So should we be putting sensitive data into the model? Should we be opening this this application up to p I... PHI, HIPAA, sort of related documentation that they should or should not be able to see? So we help with that as well. And then Jeff and the VAST team are gonna help you then build that infrastructure and build that model from a legacy compartmentalized globally dispersed infrastructure into a clean viewpoint that's built for the next generation of AI. Okay? So that's what you'll see today. I'm really excited to talk to you about it. And Jeff's gonna get into it a little bit more about why it matters. The vast story is all around AI. And when we looked even just twelve months ago, some cases six months ago, a lot of customers were at step one. Right? Which is really understanding and exploring how AI might work. So thinking about the training and the model data that's out there, early pilots, a lot of customers trying to use chatbots just to see if they could read the data and be useful in that case. And what we found in talking to them around the protection scheme was they need to be protected. The reality was all the data was replicated in some way, could be rebuilt if they had to. They needed it to be fast. Right? And that's where the vast story comes in in terms of our all flash, infrastructure, but they didn't really care so much about the data. But as you look at it and as you see these kind of the steps here on on the slide, the value of the data changes and you start to operationalize that information, you need to start thinking about ways to protect that more efficiently. Right? And that's where Rubrik and NCD comes in. Because unlike the storage platform, which does allow snapshots and replication, you need to be able to make sure the data is protected in different ways. Right? Immutability, air gapped, and be able to test that recovery because that data starts to actually have operational value to it. Right? If you're like me, you started to use Claude or Gemini or any other ones that are out there. And as you use more and more of that, what you build in that information repository is important to you. But now think about it at the enterprise scale, the information that's in there is much more valuable to that company over time. And then finally, once your business starts to use these things more efficiently, the agents are now running, continually operationalizing your your workforce there and improving the overall, functionality and, efficiencies of things. Now you've got a whole different mission behind that data. Right? That data is now mission critical, and the idea is I need to be able to make sure I can recover that as fast as possible because I can't afford to be down if I lose that information. I also may have new or yet undetermined compliance requirements around that data. And I can't say to someone, I'll rebuild it over time. Because the reality is where you were at step one and where you are at step three are two very different requirements from the data protection. And that's where, you know, our our partnership really comes in with with Rubik here because we recognize that we can't do it all from the vast side of things. We need our partners to help with this. And as we start digging into it, what we're finding is most customers are treating their data. It's in step two. They're backing up or protecting it, like step one, and they haven't fully understood the value and the criticality of that data that's limiting their AI system. And that is, generally speaking, unstructured information. So as we look at the environments, we recognize a lot of customers might have moved their data and are evolved into the mission critical state, but they're still protecting it as if they were exploring. They haven't actually taken the necessary steps to secure that data. And the reality is, if you're using out of date technology to actually run your AI systems and to protect your AI systems, there's no way you can actually protect all the data growth that's happening in these systems. So as I mentioned, AI data is generally unstructured information. It's files, it's objects, it's videos. There are things in your AI system or more importantly in your broader, infrastructure environment that generally isn't protected correctly. Right? So you can see here on the slide, 92 of enterprise data is unstructured, yet 85% of it is never regularly backed up. Now regularly backed up generally means over and over again. It may mean once a month for some environments. It may mean once a year. But, ideally, as your data changes, you wanna make sure you've got the current version ever protected. In a lot of cases, I've seen in the past. Right? You're talking about protecting two or three or four petabytes of unstructured data. It might take you several days or weeks to back that thing up using legacy systems or legacy solutions. The problem with that is, obviously, as your data changes every day, if you're not getting good copies of it, you can't possibly expect to recover to the most current version, right, or the version you were working on yesterday. And as the growth in the environment expands, right, you see here, we're expecting a seven x growth in unstructured data largely driven by AI workloads. If you don't figure out how... Way to modernize your solutions, you will never be able to actually protect it correctly. Therefore, you'll never be able to recover it correctly when need be. And as you can see, you know, 93% of the data being unstructured, if it's 10 times the cost of downtime, what's really happening in this situation is those bad actors know your critical data lives in these unstructured systems. They know if they go after those first, you are gonna be responsible for actually figuring out how to get that data back. And likely, if you haven't protected it because it takes three days to do a backup, you probably don't have any choice other than pay the ransom. Right? So they know if they go after that, there's a definite target there, and you as the owner of that information had to figure out, okay, what can I do today? My NDMP backups of my old systems are not gonna be sufficient going forward, especially as growth continues to grow and... Or the data continues to grow and grow and grow. Right? So if you think about kind of the legacy infrastructure systems, a little bit more this in a couple slides, they aren't designed for this large scale build out. Right? You get the two or three petabytes of data. Okay. But I'll tell you right now, we have a vast cluster, right, running file and object together that's an exabyte in size today. There's not a system out there from five years ago that could back that up. Even NCD, we have to look at and go, okay. We recognize an exabyte of data is a lot. Probably back up the whole thing. But working together, we can figure out what's the right way to make sure we get back up and protect that mission critical data and really drive that modernization across the board to help those who are working on AI and also making sure we have the right protection in place for that. And now, to to Jeff's point, what we strive for here at Rubrik and what, my team does and what we do every day is is trying to find and help you as the customer identify what workloads, what data points, and what applications are driving the most critical data to your business that you cannot afford an extended downtime to. So we've got some verticals up on up up on here, you know, medical imaging, What happens if the PAC system goes down? What happens if we lose access to radiology? What happens if we lose, study to research... Cancer research? Like, what is the impact to that organization? What is the impact to that, you know, facility? And and how long is your downtime if you're not accurately protecting it up. You... We talk about oil and gas and airlines, financial records, anything with the regulation or or a sensitive data tag to it usually has some sort of security policy or or protocol tied to that, and that could cause your organization to buy heavy fines to the regulators. So my challenge to you, all listening, is if you don't have a way to determine what's critical in your business, really, really take a step back. Rubrik can help you. VAST can help you, and some other tools out there in the market can help you then identify what that criticality level is to then ensure that we have an air gap immutable daily backup done to make sure that it is protected and survives any kind of disaster or cyber event that may hit the market. Right? You as the data owner knows how important the data is to the business, and that's what we're striving to protect. We don't need to protect things that, you know, I I go on grok every now and again and create images, for fun with my friends. Right? That doesn't need to be protected. But if you have core IP and you're in engineering practice and that you are building patents and building financially tied to, datasets, like, we need to ensure that that's being protected and being stored on the right place. It's not really about backing it all up. We'd love to sell you at all. We'd love to back up that exabyte that Jeff talked about. But really where we see value for your organization is identifying and then ensuring that that's protected with the RTO and the RPO that's important to your business. So as I mentioned, when I joined about five years ago, VAST was... Our story was universal storage platform. And what you see here is the concept of a single cluster or single system that was designed to replace all the traditional tiers of storage. So your high performance flash at the top all the way down to your archive at the bottom. And we do that with an all flash system that includes data reduction and the ability to scale independently in terms of your performance and your capacity. The interesting thing is, as we've evolved our story, this picture still makes sense because now it applies to AI platforms. So there... If you think about an AI platform with your vector databases, your datasets, your compute function, wouldn't it be really easy or great if all those things were in a single platform? And that's where the vast AIOS comes in. Right? Is a single solution designed to solve what is disparate systems and problems or or components of an AI platform today. But from a storage perspective, it still holds true, which is why you can actually take the same system that is doing tremendous workloads for AI, can also be a backup target for NCD. Because of that data reduction and our ability to use QLC Flash, which is the low cost Flash, we have capabilities that other platforms out there today don't have. We get a little bit into those in terms of, why and how in the next couple of slides. So what I wanted to make sure I covered here with you guys is the kind of current architecture of VAST versus the legacy infrastructure or legacy unstructured data systems that are out there. Right? Systems that were designed for file shares have a very different build and architecture than VAST does today. So if you look at it, the... They shared nothing systems. Right? The the reality of a two controller, array with a bunch of storage behind that, they don't scale efficiently because of all the east west traffic that happens. Right? So each node owns a certain amount of the data. And when you make a call, if you're looking for some some sets of data for the model, looking for some training data, if it's across all those different controllers, it has to go and find all that data, creating latency and can actually slowing the whole system down. When we talk about AI, one of the biggest things people are trying to drive is the efficiency of the GPUs. Right? An idle GPU has lost money. So if we're making sure we can serve the data up quickly, the GPUs are satisfied, and they're actually making money on those type of things. With these legacy systems that share nothing, there is significant latency associated with it, and there's a scale problem. Right? As you scale those systems out, they become less resilient. Right? Single points of failure in this system lose access to the data. Now you have to wait for that system to rebuild in order before you can access that data again. So all of those things become challenges. And finally, the last piece of it is if you look at a lot of the flash arrays that are out there, they're not designed for low cost flash. They're designed for expensive flash. And the problem is that you guys, I'm sure, are aware in the the supply chain today, flash in general is hard to get to. And if you are reliant on a much much smaller subset of high end high performance flash, you have a challenge getting the actual system deployed. The vast world, we've taken architecture called days, which is disaggregated shared everything. And the idea here is rather than a system of controllers and disks that scale together. Right? Traditionally, with a legacy system, if you needed to buy more capacity, you buy more controllers and capacity together. If you needed more throughput, you buy more capacity and more controllers together. In the... This aggregated shared everything model, our controllers, known as c boxes at the top of the line here, and our d boxes, which are just a bunch of flash at the bottom line here, can scale independently. So if you need more performance, add more c boxes. Need more capacity, add more d boxes. And the interesting thing is because those c boxes are connected to the d boxes via an NVMe fabric, every controller sees every disk and can access all that data independently. What that allows us to do is actually do, for instance, online upgrades. I can take a controller offline, upgrade it, bring it back online with no impact to the actual availability of the data. That gives us a much higher resilience standing. Right? So we have five nines availability and actually gets better as we scale up, more controllers, more capacity. We actually get more resilient along that, But it gives us a lot of flexibility as well. Right? I can add that performance independently. I can add that capacity independently. So I can have what's, you know, an environment. I talked about that exabyte size environment. It might be 500 c nodes and a thousand d boxes. If you go the other way, it could be a thousand c nodes and it could be 500 d boxes. It could also be much smaller and much more reasonable for a lot of customers, which might be a two by three. Right? So that gives us performance and capacity that meets the customer's requirements. And all of that connectivity is actually going through to affordable off Flash. So we use QLC Flash as our standard, which is, as you guys might know, kind of the lower end Flash. It typically wouldn't be used in enterprise class environment because of our architecture and the way we actually write the data. We're able to extend the life of those QLCs and make them enterprise great, as part of our system. If I didn't talk a little bit about our native capabilities, I'd be a little bit remiss. But I also understand you, we're here talking about NCD and VAST for a reason. So like all storage platforms, we have Snapshot capabilities and we have replication capabilities. Right? We have many to one, one to many. The idea there is we have that standard protection that customers expect a platform to have. The challenge is, and we've talked a little bit about here, it's not air gapped. Right. I don't have a point in time capability to be able to roll back and find individual files. It's really good for Doctor purposes. But for the purposes and the use cases that we're talking about here, Snapshots don't quite hit the mark, my opinion on things. But I know that if I needed to recover entire volume, I could roll back Snapshot. I will tell you after trying it myself, the interface may not be the easiest to use. Right? So that's where using a solution like NCD, which Ricky's gonna walk us through here in a little bit, gives a lot more capability, a lot more granularity in terms of what capabilities exist in the environment. I will also point out though that we can actually replicate the data. We do this oftentimes for those large scale environments where the customer may not necessarily know where all the critical data is, but they know they wanna have a copy of it. Okay. But now, once we actually understand what's that critical data we wanna protect, there are much better solutions out there than doing Snap and Replication. The whole point of this conversation is to talk about where VAST and NCD work well together. Right? And we're talking about building a cyber resilience solution here to make sure that unstructured data, whether it's AI or file data, is protected sufficiently. Right? So I've talked a lot about AI with VAST, and clearly that is where we wanna be in the market. But we're also an unstructured data platform. We're able to store file object, SMB data, all in the same platform. We actually can store block data as well. And the reality is customers need a way to protect that. If you look at some of the other solutions that are out there, they tend to be legacy based. They're single threaded. And we had some challenges in terms of being able to back up large protect... Large environments working that way. So working with Ricky and his team, we landed on NCD as as the solution to protect those vast workloads. As you look at this, and there's a lot of detail on the slide, I apologize, but on the left hand side you... Of the of the image here, you see vast as a source. NCD in the middle is the data mover and is the solution of, the cyber resilient component of things. And on the right hand side, ideally, since I do work for the company, I'd like you to send it to another vast cluster. But the reality is... And we're gonna talk a little bit more about this. NCD has lots of flexibility and capability in terms of where that data can go. And, again, it's part of the overall solution to be able to say to the customers and to our customers, how can I provide the right solution here? Make sure I have an immutability, for the data. I have an air gap copy, and I can recover as quickly as possible. From the vast perspective, where we love NCD, is there's two components of things that we like. Right? So the idea that it can actually discover data, but also can help us migrate data. Right? So one of the use cases that we know we look at in the AI world is how can we get data from an environment into the VAST clusters for AI purposes? That's where NCD really does come in. And Ricky's gonna go through the use cases here a little more detail, and I look forward to hearing what he has to talk about. Thanks, Jeff, for that. And and, really, the first use case I'm gonna talk about today is is what we've been... We've been talking about here for the last few minutes. But, legacy backup stack to the unified cyber resilient architecture. Okay? So a lot of you as organizations are probably probably look a lot like you do on the left. Think of a legacy backup stack. You have your applications up top. You have NAS storage. It could be NetApp. It could be Dell. It could be VAST. It could be Qumulo, Pure. You name it. And then you're layering on top of that all different tools and products to achieve a backup. So you've got your backup software. It could be a competitor of Rubrics, and it could be landing on maybe even a different, you know, s three target or a dedupe appliance that that sits on a landing zone for the backups. I have some of the most technically advanced organizations that we work with with petabytes and petabytes and petabytes of data, billions of files that still use tape. Right? So what what does that give you and what do you struggle with? Right? You have multiple tools, multiple points of failure. Some of our competitors architecture are still based on, you know, hypervisor VMs and agents and nodes. Some customers still use NDMP. Some customers I've talked to, I've talked to very large organizations that have a single copy or just rely on snapshots. It's all okay for the time being, but I'm I'm here to tell you there's a better way. Right? And if you just rely on snapshots, I'm not here to convince you not to do snapshots with VAST. I wanna be very clear. Every storage, every piece of data should have a secondary copy to a snapshot that's replicated to a different data center for business continuity and d r. Right? We're not here to replace that. You should absolutely have that. But with all these different silos in your legacy backup stack, it's very, very complex to manage. And you, as a a data administrator, are probably struggling with meeting backup windows. You're probably getting questions from the board or your leadership of what the recovery looks like or how long does it have to take and probably unproven or, you know, untested really at this point. We see it all and we talk to all different customers of shapes and sizes. Really, what we're here to talk to you about is with your investment in VAST for a, next level infrastructure platform. We have a direct partnership with VAST. We work with Jeff and his team, and we built a a purpose built API for VAST that helps us take snapshots of the VAST data on the storage. And then we use that snapshot to facilitate a backup with NAS CloudDirect, which which I'll talk about here in a second. But then we delete that snapshot once we're done with it. And, really, what NCD does really well... Think of it as a data mover at its core. It's really, really good at scanning, indexing, and moving data from source and vast in this scenario and then to target and target of your choice. Right? We'd love to sell more vast as that target like Jeff mentioned, but we have customers that that write to a hyperscaler of your choice and the tier of the hyperscaler of your choice. We have no limits or, requirements like some other backups tool that it has to land on s three standard. It could rain on directly to Glacier and to retrieval. So we give you flexibility. Why do we give you that flexibility? Because it's important to have the right RTO and the right cost profile for that application and that workload. And then, secondly on that, backup's great. We're gonna help you meet your backup window. We do really well at large objects, small objects. We have 100 petabyte shops, 50 petabyte shops, 100 terabyte shops. It doesn't matter to us. We can scale our data mover, our our NAS Cloud direct virtual machine. We can scale that up and down to meet the throughput and the, really the throughput that your organization requires to facilitate that backup. Alright? But once we facilitate that backup in our backup format, it's gonna be immutable. It's gonna be air gapped. It's gonna have all the great security tools that Rubik prides itself on, threat hunting, threat monitoring, threat detection, sensitive data classification. But the the the most important thing is recovery. Okay? So we can we can recover singular files. We have, like, a search like engine that you could search for specific files within your cluster and your fleet to recover that same that same file that you you you know. And then, like, with our threat analytics, we can scan these files on the backup and then determine, hey, malicious activity got in at this time. Let's recover from the last snapshot that we took. But we give you flexibility to where you wanna write that data back to. So in a situation where maybe your source system is compromised, maybe somebody gets in, elevates permission, locks you out of that source system, we could recover to this, a a different target that speaks the same protocol of the data that we protected. All of this is tied into Rubrik Security Cloud, which is our management plan if you're not familiar with Rubrik. But that's where we facilitate all of our our product portfolio backups. So everything from VMs on prem, SQL databases, SaaS workloads, cloud workloads, all of those are under Rubrik's Security Cloud. And we believe in SLA policies and as much automation as you can have from from a backup company. But, we create SLAs, we call them, and think of those as backup jobs. And we can do multiple backup jobs. So you could send one, maybe thirty day backup to a hot tier within AWS or onto VAST, and then maybe send a second or your weeklies and monthlies to an archival source to ensure that you meet the RTO of your business. So all of this is built in, and it's really, really good at doing large, unstructured datasets. That's our bread and butter because of the scalability of that VM and how quickly we we scan, index, and move the data. We're talking directly to the protocols themselves. We've written that. So we're very, very fast at what we do, and we can scale to meet your needs. Anything I missed, Jeff? One of the things that's important when I talk about my customers. Right? Typically, VAST is being deployed net new in an environment, and a lot of our unstructured data customers have other platforms out there. And that's where NCD really does make a difference. As I mentioned, we do have snapshots. They all have snapshots. But managing those systems at snapshot level doesn't work real well. You have a single pane of glass, a single management infrastructure, and it's tied into maybe your other data protection like Rubrik you you mentioned there. It actually gives the users a lot more visibility and a lot more manageability, to be able to actually find and protect that data correctly and recover it when needed. Yeah. I really appreciate that. And then, the second use case, Jeff mentioned it, teed it up for me perfectly. Since we are, really really good at scanning, indexing, and moving data, One of our, key use cases that we help customers with is the mobility of the set date... Data. We have a lot of customers that have a multi vendor fleet within their, storage environment, And we have, the the tool built in, we call it copy, that allows us to move data from source, you know, SMB share to, NetApp, maybe to an SMB share on VaaS. So we facilitate migrations and movement of the data for a like for like protocol while maintaining those permissions or ACLs with the move. Since we are built for speed and we talk directly to the protocol, we can achieve some really fast move... Movement when we we decide that data. This is built in. And, really, what does that give you? That that gives you vendor agnostic mobility. We have some customers that, customer of ours migrated a 100 petabytes of GCP data to AWS, and we facilitated that move. We have some customers that, some customers that, you know, have different workloads where they're using all flash for for key applications. They're processing on it. They're working on it. And then after it's done and it's through its cycle, they'll... They can move it over to maybe spinning disk. So we give you a lot of flexibility. In our partnership with VAST, we help accelerate the movement of customers off legacy storage systems to VAST's platform, and we do that really well in a partnership with VAST. So if you're in a situation where you're very interested in VAST and you you... You're worried about your... Maybe your file servers, your legacy storage, something on the other side, and you're worried about that migration itself, we can come in and help you and help you move quick, quickly to VAST, get the data on VAST, and get you making intelligent decisions with that data on VAST as quick as possible. It also mentions on the slide, we do data classification. Sensitive data classification is what we call it. It allows us to scan for sensitivity in the backup. So we have prebuilt engines. We'll look through that data and then classify that data based on prebuilt tags and triggers within our fleet to scan for things like PII, PHI, financial data. Think about if you're searching for for files or shares that have, Social Security numbers. Right? We can pull that information out. We can help you look for your critical IP within your fleet. We can do a lot of that. So all of these things are built in within Rubrik and with a partnership with VAST to help you make data management decisions, but also make sure that you have sensitive data, classified data, actually trained or stolen. So one of the interesting things to add to that is when data lands on VAST, whether it's via the migration tools we're talking about here or it's been classified via NCD. VAST has triggers built into the system that allow us to actually look at that data and go, okay. What do we wanna do with that now? Oh, this is part of our AI platform. Great. Let's vectorize it and embed it so that as the data is being moved from your legacy platforms, it's AI ready as soon as it hits the vast platform. And using the classification engine there to make sure the data that go... Going into the vast system is clean to use. Right? And they make sure that we actually have the right access to that information, which is part of the great usage of the story. Right? Yes. It can migrate data, consolidate data from legacy infrastructure to VAST. But if you're thinking about your VAST cluster as being an AI platform, it has advanced capabilities here that combined allows you to actually get that data getting ready immediately as it lands. Yeah. Great point, Jeff. Thank you. The last one... And and again, backups are great. Rubrik's a a cyber resilient company. We we talk a lot about backups, but really the bread and butter is in recovery of it. Okay? So we talk a lot about Rubrik of your minimal viable business. Have you as an organization took the time to categorize and organize what are the key applications, what are the key services that you provide your end users that if something dire was to happen to your organization that you needed to come back first. So, the tools that we have within our fleet all add together in our partnerships with VAST and companies like CrowdStrike, with Google Mandiant, with Palo Alto, and all of those other security vendors, all tied together to ensure that when the bad day happens, it's not if, but but truly when, can I recover and can I recover to the needs that fit my business? And that's what we do really, really well. So we've have recovery testing built in. We we do allow... We have a ransomware recovery team that's prebuilt for customers to, take advantage of. These are a group of experts in security and experts in rubric that help get our customers back quicker. And, Jeff, do you wanna talk a little bit about the, the data reduction on VAST? Yeah. Absolutely. As part of our platform, we have data deduplication. We call it similarity data reduction. And what that allows us to do is treat recovery very differently than other platforms that are out there. So as Ricky mentioned, right, when NCD does a backup, it's written in NCD format. But because we're actually able to deduplicate and reduce data at the byte level, we're actually able to recognize the data in a way that allows us to then deduplicate it for recovery purposes. And what I mean by that is you can see in the example here, the backup of the source data through NCD to an s three bucket, I can then create an NFS data store on the vast cluster that was my backup target. Now this is an all flash system. So in some cases, it's gonna be more performant than your your primary platform out there, your production environment. So now I can create an NFS data store, present that data store out, and run the NCD restore directly to that to that data store itself. And what that does, you can see the example here, and and the math's a little bit off. But, with the example that we test in the in the Rubrik Lab was we did 10 terabytes of line structure data recovered back to the platform where it was originally backed up to, only consumed 800 gigabytes of capacity. So you're effectively getting a high performance, in this case, NFS data store on your backup target for very little storage capacity. And where the 2% actually comes from, we actually... Across all the testing we did, it varied depending on the dataset between 2% overhead and 10% overhead, depending on how much data it changed and what was going on in the environment at that point in time. But what that allows you to do is then present that NFS data store out to production. The most likely scenario in this case is there's there's two really. One is I'm testing. If you wanna be able to test your recovery capabilities and make sure you can access that data, but don't wanna impact your production environment, great. Run the restore test here. But more likely, you'll receive more often than not is that production environment's unavailable. Right? When an attack happens, generally, those things get locked down. They don't wanna have them exposed to anything else. You cannot use those systems anymore, but you still need to access your data. Well, great. Here's an alternative that allows you to present that data immediately once the resource complete that, have a full environment up and running and can access that data right away. Yeah. That's a great point, Jeff. And it is... And I'm sure a lot of customers that are listening today. Right? It's a truly a TCO conversation as well. Right? The amount of storage space that we can save and and and a partnership with Rubrik and VAST can save you, it's it's quite phenomenal in terms of costs. Appreciate you pointing that out because you're right. We need a lot less capacity as the target because we're able to reduce the data on the backup side as well as on the recovery aspect of things. Absolutely. And, well, a few things I wanna move to next. So, it... And like any great partnership, right, we... We've spent a lot of time with our engineering teams together, working together to build out this solution. We've got some some great white papers that we can share with you. I think that will be a follow-up here as well. But we've tested this. We vetted this, and I'll talk about a customer that uses it today in the next slide. But it's fast. It's really fast. I mean, we can move a lot of data, and we can recover a lot of data really, really quick. I know Jeff just talked about the data reduction on the bottom right, but I save you money. We can we can push as fast as your physics and your network allows us to with this technology. And we're seeing a lot of benefits for customers, especially in the the multi petabyte range that have a large workload that needs to be protected within a backup window. Maybe that's nightly if you're doing dailies, we don't wanna spill over into production hours. Right? We don't wanna take any performance away from those production sources and from your end users grabbing the data. So it's really critical that we meet the needs of your business to meet your RTO. So we've tested a lot of it. You could take our word for it. You can talk to some customers that do it. But this has all been tested and vetted in a joint partnership. I think one of the interesting things is when we talk to our customers about it, who have struggled with traditional or legacy backup systems, generally, the target's always the problem. Right? Wherever you're sending that data to, in some cases, it's actually the protocol like NDMP might be the problem as well. In this joint solution, the target, if it's VAST, and NCD, will not be your bottlenecks. Let me tell you a quick story about, this joint partnership. We had a customer that, had about 26 petabytes of research data on GPFS, legacy system. They were in a position to start moving to VAST. VAST did a great job of, telling their story and selling the value to them. And they were really struggling. They were using Robocopy and rsync and some of these legacy migration tools to get the data onto VAST. It was going very, very slowly as you can imagine. We were able to talk to this customer. We were able to leverage our partnership. And, really, we were able to help them and sell them on the vision of MCD to move the data quickly to VAST to help retire that system and start leveraging their return on investment with VAST. And then, secondarily, which we talked about in the two use cases before, once the data lands on VAST, we were gonna protect it. And that's exactly what we did. So they trusted us with not only the movement of the data, which we all know a migration is not everybody's favorite thing to do. We were able to speed up that migration. I think we were pushing anywhere from, you know, on average, 30 to 50 gigabytes per second, of data movement, which was 10 x with what they were dealing with today, doing manual scripting and moving that data. And then we protected some other core critical lifeblood of their business, and they truly believe the cure for cancer is in this research. So it was a mission that we were really proud of and that we helped protect and we continue to protect today, but it's the scale. They're gonna continue to grow. This this data is not gonna get smaller. More research, more funding is gonna come in, and they entrusted us to keep their growing fleet, protected and migrated onto a modern infrastructure like VAST. We had a customer quote one of the administrators that we're working with. They they came to us and said, we passed this. This is technically viable. It works. It meets our needs. Now you need to go make sure that our leadership buys it. So we tested it. We validated it. We're happy to do POCs with our partners at VAST and with Rubrik. But this this is the critical data on the workloads that we wanna get behind. From the VAST perspective, clearly getting data into the system is an important component of us because you're paying for the solution. If it takes you weeks or months to move your data in, what's the value there? Right? So NCD clearly helps accelerate that. I like the point you made around, they believe the answer to cancer is in this data. Right? And that is, for us, the important part. Right? It's not just, like, music files. It's actually critical data that once it's on an all flash system, they can get to the data more efficiently and understand it and actually run their models much more efficiently as well. And getting it into that system as fast as possible is the key. Thanks, Jeff. That's it for us today. Really appreciate you spent some time with us. We hope you learned something. We hope you got some valuable insights. We work a lot with our friends at VAST. We have a lot of joint customers. We're happy to get you in touch with them. But, really, what my challenge would be is call myself, call Jeff, call your local Rubrik or VAST teams if you're aware. If you're not, give us your info. We'll sync you up. But see it in action. We can we can provide a demo with you and your use case specifically. We encourage at Rubrik to test if it makes sense for both organizations to do so. We're happy to show you and and talk more about your organization. We'll have a few giveaways in in terms of, technical documentation about the joint partnership. We encourage you to read those. But if you're evaluating storage, if you're evaluating backup and recovery today, there's no better place than to look that than Rubrik and VAST. So thank you.