HPE Data Services with Sandeep Singh

Machine-generated from the episode audio and not hand-corrected, so names and technical terms may be imperfect. The audio is authoritative.

All right, you're listening to yet another episode of the CTO Advisor, a long-time listener, and I want to say guest, but I don't know if I've ever had Sandeep Singh, who leads the product management for infrastructure at HPE, and also, interestingly enough, CMO for Zerto. We'll get into that in a minute. Sandeep, have you been on the podcast? I know you've been on the CTO Advisor, CTO Dose, but I don't know if we've had you on audio before. Yeah, Keith. First of all, thank you for having me.

Welcome to the audience. I think we had a chance to sit down and do a podcast a couple of years back at VMworld, so it's been a while, and it's great to be back in that sense. All right, so typically, when I've had you on, it's been sponsored content. This isn't a sponsored show, so we'll get the classic CTO Advisor and HPE conversation in which we're taking an industry-broad look at the challenges that's surrounding enterprise IT, storage services, et cetera. Then we'll get into the HPE conversation specifically.

So, Sandeep, with that said, as I look out at the landscape, what do you see as the number one challenge facing enterprises when it comes to their, let's talk big picture, data services themselves? Yeah, great question, Keith. I actually see two major shifts in the market that's taking place. First is what I call this notion of cloud first to cloud anywhere or cloud everywhere. That's a pivot that we're seeing in the market where CIOs, where their teams have taken workloads over to the public cloud and seen the agility that's unlocked with the cloud operational experience.

Now they're stepping back and asking basically, how can I get this agility everywhere where my data and workloads live? And that's that shift from cloud first to cloud everywhere. The second key shift is really data first modernization or cloud first to data first, which is that data has truly become the lifeblood of every organization. I think we all recognize that. When you think about data, the journey starts with ingesting it, storing it, and serving it up to the applications. But then you have to really think about how do I protect it from disruptions and ransomware and cyber threats, which have become the net new reality, how do I back it up?

As well as how do I ultimately analyze and uncover the insights from data? So those are the two fundamental shifts that we're seeing in the marketplace. And really the challenge comes down to the complexity that IT is still facing. That's kind of the major or the core challenge that we see there. So let's dive into this data first attitude that you're seeing within enterprises. And let me put an anecdotal experience with this. I talked to the CIO of one of the largest retailers in the nation, convenience store retailers, and they were on a modernization at modernization journey where they were doing a bunch of really cloud first initiatives, really transforming their operations and their relationships digitally, being able to meet the needs of their suppliers digitally and meet the needs of their customer and changing their customer experience to a digital first experience.

And as he kind of laid out the app portfolio, very impressive. They were using some data points from more applications from the SAS world. So Salesforce, they were rebuilding some of the data services in their data center. Very, very great. Fresh look from a cloud first perspective. Then I drew what I call my data infrastructure picture for him. And I say, you know what? How do you how do you apply the policy, the your security policy, your access control policy or your data that's living in Salesforce and your data that's living in the private data center and your data that's living in AWS or the cloud in general?

How do you ensure a consistent data policy and data protection scheme? He kind of scratched his head and said, you know what, that's a valid point. Talk to me in three or four years while I deal with like the application level transformation first. Have you run into similar stories? Well, Keith, I think that's a fantastic point, first of all, and a great insight across the board. Clearly, as customers are thinking about how do I modernize my apps, they need to be hand in hand thinking about data and how do they modernize the overall data management because data as it underpins their apps, data is also distributed across edge core and cloud.

So when customers start to think about that, they have to be thinking about this in tandem. Right. There's multiple ways of thinking through that part of it. One is obviously that for a lot of the customers, their mission critical workloads and their data are sitting on prem. So the first challenge or the question that arises is, well, geez, am I going to take this over to the public cloud to be able to modernize this? If so, what do I need to do in terms of refactoring or re-architecting?

And what risk does that introduce to get that end benefit? And alternatively, if you can bring that cloud experience right where it is on prem, that's a lower risk way of transforming in place. So that's one element of it. A different way of thinking about it is basically with data that is distributed across various clouds, as well as on prem and edge, then by very nature, what you have to build an architect for is a hybrid cloud world. And then really think through how can policies be applied globally so you can get a uniform set of policies that you can implement and monitor and be compliant with?

So I think there's this sense that, especially that gap from that CIO, like, you know what, that's a very valid question, Keith. But this seems not possible or too big to tackle. And I have to be up front. There's not a solution on the market today that allows me to take my access control. Let's, you know, zoom in and access control and and solve the access control problem to begin with, because, you know, there's the core identity problem. Where is how do I say that the data that's in Salesforce is the same data that's on prem or the same classification of the data?

I can't. And there's no centralized policy management. So that problem, I think, is intractable right now. There isn't a a solution on the market. But as we take data, let's zero in and data protection and put your kind of your Zerto cap on for a sec. And not just Zerto. You also lead. Is it fair to say that you also lead the data backup solution on the traditional HP software side? It's I'm part of the storage business unit and within the storage business unit, we are offering the overall backup as a service.

That's the HP backup and recovery service. So it's fair to talk about that. So I guess the tough question is, how does the CEO take advantage of that? Like the let's say that he wants to bypass the VP of storage. And there's this rich data platform that the VP of storage has enabled, but there's no communication between the two. How does the CEO kind of self service his needs to do exactly that? Yeah, that's another fantastic point. The it starts with having this platform through HPE GreenLake to be able to get access through HPE GreenLake.

There is role based access that's been put into place so that different roles, as you think about the administrative boundaries across an organization can have access based on their type of roles. So, for example, in this CEO role, a specific role based access can be granted that then gives them access to their data, for example. And as we continue to build out the GreenLake platform for storage and data protection and data visibility, then that CEO will be able to act on a given data set for deriving insights.

What you also touched upon, Keith, is this notion of ultimately how can I grant and enable self-service agility? T. can be offloaded. But much more importantly, line of business owners, the database admins or the data scientists can gain self-service agility to be able to act on data and really for getting their application workloads up and running much, much faster. So. I get HPE GreenLake and I get block services in general, but when I talk to folks who are sometimes partners of HPE, they say, you know what, object storage is the future for everything, not just data analytics or cloud based applications.

They've they've said, you know what data object storage is now the primary storage. So with that type of marketing out there, how should CTOs, CIOs think of something like a block storage? Isn't that just more legacy infrastructure that CTOs no longer want? So great point. Clearly in the market, there is unstructured data, which is a very high growth market in terms of just the amount of data that is being generated. The existing applications and application workloads, the database based OLTP applications, those are historically block storage oriented.

And there is a large overall market that exists. All of the applications that exist today, those are clearly block oriented. But to your point, as we think about the unstructured data and the growth of cloud native apps, where they're being architected with a back end object storage semantics there, where those object storage semantics have just become much more favorable and a way for enabling developers to make certain assumptions about the back end and being able to scale it seamlessly. That is the growth area when we think about cloud native apps and or a variety of other use cases that are continuing to emerge.

So we will see a future where there is this coexistence of existing block storage based workloads and yet the emergence of object based storage workloads as well. So how should. People building. Infrastructures for both, you know, traditional block storage workloads and object storage based workloads, think about HPE data services and generating GreenLake, should they mainly focus on GreenLake for building these traditional apps that require NFS, iSCSI, et cetera, for the HPE GreenLake and go somewhere else for object storage or is there a solution for both camps?

Keith, last year in 2021, what we announced was that we're fundamentally transforming the entire storage business from a leader in storage to becoming a leader in overall end to end data services. And that's through the lens of the HPE GreenLake cloud platform. As part of that announcement, that blueprint or the vision that we shared are showcases how we're building the end to end the underlying cloud native data infrastructure. On top of that, we're delivering the overall infrastructure as a service, in this case, storage as a service.

So block storage as a service is that beginning. But clearly, as we fast forward and continue to deliver, that will be inclusive of, for example, object as a service there. And then the third layer building on top of that is the heterogeneous set of data services. So we're obviously on a journey, our customers are on a journey with us, and they can fully expect to see us being able to deliver and address their needs for both the structured data sets, as well as the unstructured data sets.

So, Sandy, talk to me about HPE's block storage. Like, how does HPE's block storage service enable this data, what I call data infrastructure, this ability to have this base capability that I can extend across groups and across applications and types of applications and environments? Yeah, Keith, one of the fundamental shifts that we're seeing in the marketplace is that customers are increasingly wanting to enable this notion of self-service agility, where IT has been under constant pressure from line of business owners of, I've got an application workload to deploy, I want to deploy it now, just put it in the cloud, and yet IT is challenged with the thinking about the broader considerations for the data portion of it.

And that is motivating a shift for customers to shift over to deploying storage as a service. And we actually earlier introduced block storage as a service. But fundamentally, what that is looking to enable and overcome is three key challenges that have existed in the marketplace for customers. One has been that the storage as a service offers that have been there so far really came with premium managed services, and that's fantastic for a certain set of customers, but it limits the broader adoption.

The second key area is that the existing storage as a service offers come with four nines of availability, yet we both know that the business critical mission critical workloads require six nines or 100% availability. But the third one is the most important, where customers are after that self-service agility, yet the existing storage as a service offers still come with traditional storage management that requires domain expertise. And with the introduction of HPE block storage as a service, we are ending those compromises, enabling whether it's line of business owners, database admins to gain that self-service agility.

And that's thanks to intent-based provisioning, which is powered underneath the covers by HPE InfoSight, which brings it along AIOps so that customers can just identify their application workload and be able to identify the amount of storage capacity needed. And through that, we can intelligently determine across their fleet where that workload should be provisioned. So Sandeep, I really appreciate you stopping by and being brave enough to come on to CTO Advisor in a non-sponsored thing. There was no handlers on. There was nobody saying, no, don't answer that question.

You can't ask that question. It's been a pretty free-flowing conversation about the overall data services market at large and what HPE has going on. com. You want to ask me questions and say, you know what, Keith, even in this editorial thing, you didn't go after HPE tough enough. We want to hear about Ezmeral and the entire end-to-end vision. To be fair, Sandeep is focused on data services. We'll have somebody from Ezmeral on and we'll talk Ezmeral and we'll talk big vision after that.

These conversations with large OEMs are very, very critical. I'm a HPE customer. I had these very questions as a HPE customer for Sandeep. I'll have more questions around Ezmeral versus VMware, Ezmeral with VMware, Ezmeral with I'm a Red Hat environment. All these complexities are things we're dealing with in the CTO Advisor hybrid infrastructure. Again, you want to find out more, visit us on the Web, talk to you next CTO Dose.