Episode Transcript
[00:00:15] Speaker A: Welcome to K beyond the go. This week I'm on the ground at NetApp Insight 2024 conference at the MGM grand in the heart of Las Vegas. For this bonus series, we've been lucky enough to have lined up conversations with the selection of NetApp executives and other guests exploring the future of data and Aihdenhe. Stay tuned for the Insight track from some of the world's leading authorities, presenting at Insight 2024 as KBI Media brings you all of the highlights.
Joining me now in person is Ashish Dhawan, senior vice president, general manager and chief revenue officer of cloud business unit at Netapp. And today we're discussing the drive to enterprise cloud adoption. So, Ashish, thanks for joining the Wildcath.
[00:00:57] Speaker B: My pleasure. Thank you so much for having me.
[00:00:59] Speaker A: So, Ashish, let's start with your view on what's driving enterprise cloud adoption.
[00:01:04] Speaker B: I think cloud adoption in our customers is largely to do with our customers need to innovate faster.
They also need to make sure that they have a very optimal it infrastructure. They also need to have a very elastic it infrastructure. Cloud computing really helps them achieve all these at one goal.
When they move their applications and their workloads to the cloud, they're able to command resources like compute and storage and networking at will. They're able to have infrastructure which is elastic. They're able to ensure that with the compute and storage of the hyperscaler, they're also able to use services like open source services, like data services, and now increasingly AI and Gemini services. So that's what's driving our customers to the cloud. It's making sure that their it infrastructure is optimal, elastic, and they can innovate faster.
[00:02:10] Speaker A: And do you think historically, people were sort of apprehensive about migrating to the cloud, would you say in your experience?
[00:02:16] Speaker B: Well, a few years ago, migrating the cloud seemed like a difficult proposition. But we NetApp's cloud storage solutions along which are jointly built with our hyperscalers, I've tried to make it very simple for our customers. Migrations need not be difficult, because what we try and offer to our customers is, first of all, technology that is very akin to what they used to on prem feature set, which is very similar to what they're used to on. But given in the first one P format, which our customers are used to, these services that we've built with Amazon, Microsoft and Google are first party services, dog storage services.
And these first party services are actually positioned through the hyperscalers console. The hyperscalers. They're part of the migration story of the hyperscalers they're part of each and every application migration sales play that the hyperscalers have to. Our customers are now finding it easier to migrate because they're able to complement the hyperscalers compute services along with the storage services that they've built with us, making it easier for customers to migrate their most critical workloads on them.
[00:03:40] Speaker A: So, speaking of hyperscalers, I want to now talk a little bit more on, to hear your thoughts on the perspective from your customers. What are you sort of hearing from that point of view?
[00:03:49] Speaker B: Yeah. So customers have started cloud at a large scale. We're now having the most critical assets, which is applications, databases, and their data are on the cloud. Having said that, they're also choosing the right cloud for the right purpose. And therefore for us, it was important to make sure that we develop our first party cloud storage services with each of the hyperscalers. So, for example, you're the only company that has these services with each of the hyperscalers. For example, with Amazon, we have Amazon FFXN, which is cloud storage service for AWS. With Google, we've got Google Cloud NetApp volumes, and with Azure, we've got Azure NetApp files. These services integrate into the migration plays off the hyperscalers and make it easier for our customers to migrate their critical assets on us. Now, these partnerships are critical to us because with these first party services, we are not only able to help our customers migrate their assets, but once migrated, they're able to utilize the data on the cloud to get the AI and gen AI outcomes that they so desire. The outcome that you're looking at with our customers is help them migrate first so that they can have an agile IT infrastructure, and then once they've done so, use the data that they've migrated so that they can get the AI engineering outcomes that they so desire in these days.
[00:05:25] Speaker A: And then on that, what would you say are some of the main challenges customers are facing when trying to probably adopt a cloud first approach?
[00:05:33] Speaker B: Yeah, the biggest challenge that any customer faces from a cloud for a cloudflare approach is actually top down alignment. A lot of times we see senior management wanting to adopt cloud technology, but there's no alignment with the rest of the leadership. So ensuring that there is top down alignment within the leadership of the customer is extremely important. We help our customers arrive at that. The second challenge that they have is making sure that they have enough skill sets right, making sure that the skill sets that they have know how to implement a cloud storage solution along with us, along with the hyperscale is extremely important, and therefore, we try and, you know, provide the level of training, certification and ongoing support for them to make sure that that never remains an issue. And lastly, we helped them build very good cloud economies, cost cases, so that they realized for the long term, they're actually making investment, which is not just efficient for the current, but also will pay rich dividends in the future, making their it infrastructure agile and more innovative.
[00:06:52] Speaker A: And you said before, no alignment with the leadership.
Why is that the case from your experience?
[00:06:58] Speaker B: Yeah. In many companies, one part of a company wants to, let's say, adopt cloud technology, but the other part of the company has certain assets which they're very used to. So it's a question of ensuring that they also understand the long term benefits of using cloud technology. And that's why that alignment is super pretty. It's the number one inhibitor to implementing cloud computing is making sure that there is alignment within the senior leadership of the organization. Making sure that the CEO is aligned with the CIO, who are internal, aligned with the various business units, is very critical for our inclementation of counter plans.
[00:07:47] Speaker A: So how would you approach, or what would be your strategy to getting that alignment from a leadership perspective?
[00:07:51] Speaker B: Yeah, you know, we first approach it from a cultural perspective. We look at it. Technology almost always means a little bit of a cultural overvalue.
You're asking the company not to own assets, but to rely on a utility that is provided by a cloud provider.
And that utility obviously comes with feature sets that they're not used to, SLA's that are slightly different, terms of engagement that are slightly different. So ensuring that once we are able to earn the trust of, let's say, one part of the senior leadership, that trust is, is propagated in the western leadership team. For him, we do this by explaining how these differences actually are for the better. We go to the customer and make sure that these differences are understood. And we also explain what kind of cultural differences they need to implement or they need to have in the form of different mechanisms so that they can utilize this technology better. You know, they're not now paying in for kits that they will purchase, but they're paying for services that they will consume. It's a consumption based, consumption based engagement that we have, which is very successful.
So in their success is our success, so therefore we focus more on their customer success. And that's a huge change in mindset that we try and help them understand.
[00:09:40] Speaker A: Joining me now in person is Gagan Gulati, general manager, data services at Netapp. And today we're discussing the focus on data protection and governance. So, Gargan, thanks for joining and welcome.
[00:09:49] Speaker C: Thank you very much.
[00:09:49] Speaker A: So, as we all know, data is king. Tell us a little bit more about the focus specifically on data protection and governance from your point of view.
[00:09:58] Speaker C: Yeah, absolutely. So look, our customers are in the data journey. They are hybrid today, multi cloud. Pretty much all our customers have the data on all types of clouds today.
At the same time, data is sprawling new workloads. AI is coming in at the same time, cyber threats are a big deal. Almost all our customers are at a place where they need to protect their data button, make sure that their crown jewels are saved and they can recover very quickly when it comes to resiliency against the cyber threat. So that's fine. And AI is just an example of a workload that's brought a big focus on data governance, to understand what their data is, to be able to make sense of it then, and to hand it over to the right people so they can access it to stay compliant with new regulations like GDPR and others. So generally speaking, you're right, data is the king. And at NetApp, we have a huge focus, therefore, on data prediction and governance, because we are the intelligent data infrastructure company.
[00:11:06] Speaker A: And in terms of governance, what's your sort of view on how people should be addressing that? Because, again, speaking of Jeff Baxter before you, he was sort of talking about a majority of companies don't really even know all their data is, for example. So how do people effectively govern that, would you say?
[00:11:22] Speaker C: Yeah, you're absolutely right. And I think this is the constant problem. You have to know your data before you can govern your data. You know, we talk to a lot of, we talk to a lot of storage admins, and we basically ask them simple questions, why do you want to buy more storage? Why not governing your data better and delete the old data? Well, because they don't know your data, their own data. They may have temp files sitting in a storage unit with, you know, let's assume petabytes or terabytes of stem files, which you can delete because they were ten years old, but they're afraid they don't know what the data looks like. If you go to them and say, I need access to this file share, they don't know whether they should give you access to the file share or not because they don't know their data. So knowing the data itself is a huge deal. And then, of course, you have to govern your data. This is to not only make sure that the companies meet their compliance needs.
[00:12:13] Speaker A: Sure.
[00:12:13] Speaker C: For example, if I have an m and a file share, you know, merger and acquisitions file share, should Dargan have access to merger and file acquisitions file share?
[00:12:22] Speaker B: No.
[00:12:23] Speaker C: Right. But how do you know that this is a MNA file shared? You need to be able to govern your data better, know your data to govern it better, to provide the right permissions to the right set of people, to be able to be resilient. And all of that is just so critical, especially now with the whole AI phase coming in, because AI is data hungry. You want to make sure your data scientists have access to the right data and only the data they should have access to. And you can't do that if you don't have good governance on that data.
[00:12:52] Speaker A: So can I ask, how can people get to the stage? Or are they getting to the stage where they're like, we need to understand inherently where our data is, because back in the day, you wanted to get all of the data like Cloudera and those types of days. Now it's like, oh, we don't want it anymore because of data breaches and holding too much PiI, for example.
[00:13:11] Speaker C: Yeah, I mean, most customers, there are customers who started their journey many, many years ago. You'll find a class of enterprise customer, the bigger financial customers, who are quite ahead in this journey, actually, in terms of understanding their data so they can govern it better. And they are. But I would say the majority of the customers are not. And the times of cybere attacks, in the time of cyber attacks that are happening today, in the times of AI, all customers are now struggling and trying to figure out how do they do that. So I would say most of them have just started a journey.
Customers I meet with regularly are asking us, you are the intelligent data infrastructure company. Can you please help? So we are building tools. We are giving them advice on how they should go ahead. We are connecting them with our partners and giving them our products. And a lot of our products, we are making them available for free now so that they can start their data governance journey and by just understanding what data they have. But I would say majority of the customers, very early stage.
[00:14:15] Speaker A: So you think products for free? What do you get for free here?
[00:14:18] Speaker C: Yeah, so it's an example of BlueXP classification. It's a product that NetApp has had for years. And our customers love that product, but we use the charts for it. But as we enter into this phase, where customers want to discover their data, to protect it, to govern it, you know, we see an inherent need there. And you know, we are in the business of helping our customers, helping our customers to the best we can in terms of discovering the data. So recently around, I think, four or five months ago, we made it freely available for our customers to use. They don't have to pay for it anymore. They can just go to blue XP and they can use it at a core functionality to end this point, that product to their, to their data, to the storage that's on NetApp. And this product comes back and tells them what's in it based on time, stance, based on the sensitivity of data, based on like a heat map, what's being used, what's not being used, so they can make sense of it. And we have seen the growth, an amazing growth. Like four times more customers use it today. Then they used it just four or five months ago. So it's an amazing.
[00:15:24] Speaker A: Why do you think that's the case?
[00:15:26] Speaker C: I mean, the case is, of course, one is just the rise of regulations, the rise of cyber threats and cycle. And number two is it's the ease of use. You don't have to go to your, you know, your finance team now, procurement team now to say, we want to do this. Oh, and the, and this will be only done in the next annual, you know, planning cycle. They can just start using it today by just going to blue xp. So I think we made it really easy for the customers and they're all starting their projects right now thanks to the urgency and need. But at the same time, we took away the biggest inhibitor that they have to go and purchase it separately. We made it part of the common licensees that they have and didn't it as well.
[00:16:08] Speaker A: When people are obviously doing the discovery phase, they're a little bit rattled by what comes back around, what they've got.
[00:16:14] Speaker C: You will be surprised. Yeah. Our customers who, for example, have run data classification have come back. I was given an example of, oh, my God. The only, it's a, it's a folder full of, it's a volume full of just temp files from a project that we discontinued ten years ago.
But it's taking 50 terabytes of space.
Right.
[00:16:36] Speaker A: A lot.
[00:16:37] Speaker C: Right. And there are files from an old project and nobody's touched them in ten years. And they're big image files, big video files, and they're useless to the company. But the storage admins so far were extremely hesitant to try to delete that or move them somewhere else. But now they have the ability to do all of that. The surprises are massive. Like, these guys are getting like, oh, my God, we didn't even know this existed because the people who use them have left the company.
[00:17:06] Speaker A: I was going to say exactly. Someone's left. Someone didn't know about it. Things changed. So then with blue XP, with the discovery side of it, is this continuous then, or.
[00:17:17] Speaker C: Yeah, yeah, absolutely. So the way to run this system is manifolds. A lot of the customers use the blue XP classification to just understand their data so they can migrate it, delete it, retain it, move it to the cloud, move it to, you know, other kind of cheaper storages. So that's one example. But the true value of blue classification is the continuous process of classifying data as more and more data changes.
Good example is compliance needs. Let's assume you basically say that I do not want any project x data to be sitting on this folder. I'm just giving an example. Or I'm using this data set for a Jenny eye project. And the Jenny project is all about marketing, helping marketing. So therefore, I don't want any lesson finance files to come into this data set.
So the power of blue XP classification is that it's continuous and it'll tell you what's going on all the time. So when things go wrong or the rules are not met or the regulations are not met, it'll come and tell you. I found these 50 files that do not meet the basic compliance needs. So you should do something about that.
[00:18:30] Speaker A: And are people doing things about it today, manually?
[00:18:34] Speaker C: Today, they're just reporting. It's a start. So pretty much everybody is in the report zone, which is. I just want to know very soon. Yeah, it's a great start. I think it's know your data to govern your data. Pretty much everybody is like, I just want to know my data. Give me the reports. I want to take it to my board of directors. I want to take it to my ciso. I want to take to my chief compliance officer and just show them what exists.
[00:18:55] Speaker A: Correct.
[00:18:55] Speaker C: And somebody else will make the decision or what we should do about it. But I think that the tide is starting to turn. Very soon there will be governance regulations, and more and more customers will now start enforcing the compliance regulations or other needs they may have.
[00:19:14] Speaker A: So I want to sort of flick over now.
[00:19:15] Speaker C: Gargan.
[00:19:16] Speaker A: And it's slightly changed gears. This is the cybersecurity podcast.
[00:19:19] Speaker C: Yes.
[00:19:20] Speaker A: And I want to talk more about autonomous ransomware protection and then how that works with ionic.
[00:19:26] Speaker C: Yeah, look, as you know, every, almost every company has gone through a ransomware attack. If they have not gone through, they will petition.
And when it comes to ransomware, when it comes to ransomware attacks, it's all about real time ransomware detection, because the sooner you can detect that there's an attack, the sooner you can respond to it.
[00:19:45] Speaker A: Sure.
[00:19:45] Speaker C: Right. And you want to do this. All of this has to happen on the primary storage where your data is, because if you don't do that on the primary storage, that by the time the backups are taken and you want to run your anomaly, anomalous anomaly detection over there, it may take 6 hours, 8 hours, and these attackers are gone by then, they're finished up. They've encrypted your entire drive. They're finished up your entire encrypted entire storage. And that's the keen. So real time ransomware detection is a feature that we have put in a few years ago into on tap, and it's worked really well. What it does, it is, and we call it ARP. And what ARP allows it to do is to say whenever there is an attack, you can respond. You can respond by generating an alert for your security operator. You can respond by taking the snapshot. You can respond by many, like blocking users and many other action.
What we have done recently is we have enhanced the capability for our customers to be able to more accurately detect the ransomware attacks in real time by introducing the power of AI models. And therefore we call it Aaron Pan. And what these new AI models do is that they're way more effective and accurate in detecting the attacks. Some of the studies, and we worked with a security breach company called Selabs to validate this, and they gave us a aaa rating on the first attempt because we were able to detect 100% of all encryption attacks that are, you know, are happening today with 99% recall, which means we don't really miss any attacks either. Right, and that is the power of AI models as you bring them into ARP, and we call it ARpaI, and that's now shipping in ONTAp, 960 in one. But that's not all.
It's not just enough to ship these models and let them be static, right. Because the attackers are moving at a fast pace. They are using AI to come up with new techniques. So therefore, we have to make sure that these AI models that we are putting into on tap also continue to stay update all the time. So we have built in a complete auto update pipeline for these models. Just like, you know, on your laptop right now, there's an antivirus running. You don't know when the antivirus company gives you a new signature, you're not rebooting the machine every time. They just keep you safe. So we are bringing the same consumer like technology into enterprise now, where your on tap systems will always remain safe, because we are going to keep sending the new AI models to your on tap system without you even knowing about them, if you wanted to, without upgrading the onhab ever, and without worrying about patching on tap. So that's been the most beautiful thing in the entire journey as we bring these AI models into on tap. That and call it the art. An.
[00:22:43] Speaker A: And there you have it. This is KB on the go. Stay tuned for more.
