 From around the globe, it's theCUBE, presenting Enterprise Digital Resilience on Hybrid and Multicloud, brought to you by IOTAHO. Hello everyone and welcome to our continuing series covering data automation brought to you by IOTAHO. Today we're going to look at how to ensure Enterprise resilience for Hybrid and Multicloud. Let's welcome in AJ Vajora, who's the CEO of IOTAHO. AJ, always good to see you again. Thanks for coming on. Great to be back, David, pleasure. And he's joined by Fadzi Ushawa Kunze, who is a global principal architect for financial services, the vertical of financial services that Red Hat, he's got deep experiences in that sector. Welcome Fadzi, good to see you. Thank you very much, happy to be here. Fadzi, let's start with you. Look, there are a lot of views on cloud and what it is. I wonder if you could explain to us how you think about what is a hybrid cloud and how it works. Sure, yes. So a hybrid cloud is an IT architecture that incorporates some degree of workload, portability, orchestration and management across multiple clouds. Those clouds could be private clouds or public clouds or even your own data centers. And how does it all work? It's all about secure interconnectivity and on demand allocation of resources across clouds. And separate clouds can become hybrid when they're seamlessly interconnected. And it is that interconnectivity that allows the workloads to be moved and how management can be unified and orchestration can work. And how well you have these interconnections has a direct impact on how well your hybrid cloud will work. Okay, so well Fadzi, staying with you for a minute. So in the early days of cloud, that term private cloud was thrown around a lot, but it often just meant virtualization of an on-prem system and a network connection to the public cloud. Let's bring it forward. What in your view, does a modern hybrid cloud architecture look like? Sure, so for modern hybrid clouds, we see that teams or organizations need to focus on the portability of applications across clouds. That's very important, right? And when organizations build applications, they need to build and deploy these applications as small collections of independently loosely coupled services. And then have those things run on the same operating system, which means in other words, running it on Linux everywhere and building cloud native applications and being able to manage and orchestrate these applications with platforms like Kubernetes or Reddit open shoot, for example. Okay, so that's Fadzi, that's definitely different from building a monolithic application that's fossilized and doesn't move. So what are the challenges for customers to get to that modern cloud as you've just described it? Is it skill sets? Is it the ability to leverage things like containers? What's your view there? So, I mean, from what we've seen around the industry, especially around financial services where I spend most of my time, we see that the first thing that we see is management, right? Now, because you have all these clouds and all these applications, you have a massive array of connections, of interconnections. You also have massive array of integrations, portability and resource allocations as well. And then orchestrating all those different moving pieces, things like storage networks and things like that. Those are really difficult to manage, right? That's one, so management is the first challenge. The second one is workload placement. Where do you place this cloud? How do you place this cloud native application? Do you, what do you keep on site or on-prem and what do you put in the cloud? That is the other challenge. The major one, the third one is security. Security now becomes the key challenge and concern for most customers. And we can talk about how to address that. Yeah, we're definitely going to dig into that. Let's bring AJ into the conversation. AJ, you and I have talked about this in the past. One of the big problems that virtually every company's face is data fragmentation. Talk a little bit about how IOTAHO unifies data across both traditional systems, legacy systems, and it connects to these modern IT environments. Yeah, a short day. I mean, Fadsay just nailed it there. It used to be about data, the volume of data and the different types of data, but as applications become more connected and interconnected, the location of that data really matters how we serve that data up to those apps. So working with Red Hat and our partnership with Red Hat being able to inject our data discovery, machine learning into these multiple different locations, whether it be an AWS or an IBM cloud or a GCP or on-prem, being able to automate that discovery and pulling that single view of where is all my data, then allows the CIO to manage costs that can do things like, one, I keep the data where it is on-premise or in my Oracle cloud or in my IBM cloud and connect the application that needs to feed off that data. And the way in which we do that is machine learning that learns over time as it recognizes different types of data, applies policies to classify that data and brings that all together with automation. Right, and that can be one of the big themes that we've talked about this on earlier episodes is really simplification, really abstracting a lot of that heavy lifting away so we can focus on things, AJ, as you just mentioned. I mean, Fadzi, one of the big challenges that, of course, we all talk about is governance across these disparate data sets. I'm curious as your thoughts, how does Red Hat really think about helping customers adhere to corporate edicts and compliance regulations, which of course are particularly acute within financial services? Oh, yeah, yeah. So for banks and payment providers, like you've just mentioned there, insurers and many other financial services firms, they have to adhere to standards such as say PCI DSS and in Europe you've got the GDPR, which requires stringent tracking, reporting, documentation, and for them to remain in compliance. And the way we recommend our customers to address these challenges is by having an automation strategy, right? And that type of strategy can help you to improve the security and compliance of your organization and reduce the risk after the business, right? And we help organizations build security and compliance from the start with our consulting services, residencies. We also offer courses that help customers to understand how to address some of these challenges. And thus also we help organizations build security into their applications with our open source middleware, middleware offerings and even using a platform like OpenShift because it allows you to run legacy applications and also containerize applications in a unified platform, right? And also that provides you with the automation and the tooling that you need to continuously monitor, manage and automate the systems for security and compliance purposes. AJ, any color you could add to this conversation? Yeah, I'm pleased to have brought up OpenShift. I mean, we're using OpenShift to be able to take that security application of controls to the data level and it's all about context. So understanding what data is there being able to assess it to say who should have access to it, which application permission should be applied to it. That's a great combination of Red Hat and Airtaho. Patsy, what about multi-cloud? Doesn't that complicate the situation even further? Maybe you could talk about some of the best practices to apply automation across not only hybrid cloud but multi-cloud as well. Yeah, sure, yeah. So the right automation solution can be the difference between cultivating an automated enterprise or automation chaos. And some of the recommendations we give our clients is to look for an automation platform that can offer the first thing is complete support. That means have an automation solution that provides, promotes IT availability and reliability with your platform so that you can provide enterprise great support including security and testing integration and clear roadmaps. The second thing is vendor interoperability in that you are going to be integrating multiple clouds. So you're going to need a solution that can connect to multiple clouds seamlessly, right? And with that comes the challenge of maintainability. So you're going to need to look into a automation solution that is easy to learn or has an easy learning curve. And then the fourth idea that we tell our customers is scalability. In the hybrid cloud space, scale is the big, big deal here. And you need to deploy an automation solution that can span across the whole enterprise in a consistent manner, right? And then also that allows you finally to integrate the multiple data centers that you have. So AJ, I mean, this is a complicated situation if a customer has to make sure things work on AWS or Azure or Google, they're going to spend all their time doing that. What can you add really to simplify that multi-cloud and hybrid cloud equation? Yeah, I can give a few customer examples here one being a manufacturer that we've worked with to drive that simplification and the real bonuses for them has been a reduction in cost. We worked with them late last year to bring the cost spend down by $10 million in 2021 so that they could hit that reduced budget. And what we brought to that was the ability to deploy using OpenShift templates into their different environments, whether it was on-premise and or in, as you mentioned, AWS, they had GCP as well for their marketing team. And across those different platforms, being able to use a template, use pre-built scripts to get up and running and catalog and discover that data within minutes, it takes away the legacy of having teams of people having to jump on workshop calls. And I know we're all on a lot of teams, Zoom calls in these current times. They're just simply isn't enough hours in the day to manually perform all of this. So, yeah, working with Red Hat, applying machine learning into those templates, those little recipes that we can put that automation to work regardless of which location the data's in allows us to pull that unified view together. Great, thank you. Fazia, I want to come back to you. So the early days of cloud, you're in the big Apple, you know, financial services really well. Cloud was like an evil word and within financial services. And obviously that's changed, it's evolved. We talk about the pandemic has even accelerated that. And when you really, you know, dug into it, when you talk to customers about their experiences with security in the cloud, it was not that it wasn't good, it was great, whatever, but it was different. And there's always this issue of skill, lack of skills and multiple tools, SecOps teams are really overburdened. But then the cloud requires, you know, new thinking. You've got the shared responsibility model. You've got to obviously have specific corporate, you know, requirements and compliance. So this is even more complicated when you introduce multiple clouds. So what are the differences that you can share from your experiences running on a sort of either on-prem or on a monocloud or, you know, and versus across clouds. What do you suggest there? Sure, you know, because of these complexities that you've explained here, misconfigurations and inadequate change control are the top security threats. So human error is what we want to avoid because as, you know, as your clouds grow with complexity, and you put humans in the mix, then the rate of errors is going to increase and that is going to expose you to security threats. So this is when automation comes in because automation will streamline and increase the consistency of your infrastructure management, also application development and even security operations to improve and your protection, compliance and change control. So you want to consistently configure resources according to pre-approved policies and you want to proactively maintain them in a repeatable fashion over the whole life cycle. And then you also want to rapid the identified system that require patches and reconfiguration and automate that process of patching and reconfiguring so that you don't have humans doing this type of thing, right? And you want to be able to easily apply patches and change our system settings according to a pre-defined baseline like I explained before in, you know, with the pre-approved policies and also you want ease of auditing and troubleshooting, right? And from a radar perspective, we provide tools that enable you to do this. We have, for example, a tool called Ansible that enables you to automate data center operations and security and also deployment of applications and also office shift itself, you know, automates most of these things and obstructs the human beings from putting their fingers and causing, you know, potentially introducing errors, right? Now in looking into the, you know, new world of multiple clouds and so forth, the differences that we're seeing here between running a single cloud or on-prem is three main areas, which is control, security and compliance, right? Control here, it means if you're on-premise or you have one cloud, you know, in most cases you have control over your data and your applications, especially if you're on-prem. However, if you're in the public cloud, there is a difference. They're the ownership, it is still yours, but your resources are running on somebody else's or the public clouds, EWS and so forth infrastructure. So people that are going to do this need to really, especially banks and governments, need to be aware of the regulatory constraints of running those applications in the public cloud. And we also help customers rationalize some of these choices. And also on security, you will see that if you're running on-premises or in a single cloud, you have more control, especially if you're on-prem, you can control the sensitive information that you have. However, in the cloud, that's a different situation, especially from personal information of employees and things like that. You need to be really careful of that. And also, again, we help you rationalize some of those choices. And then the last one is compliance. As well, you see that if you're running on-prem or in single cloud, regulations come into play again, right? And if you're running on-prem, you have control over that. You can document everything. You have access to everything that you need. But if you're going to go to the public cloud, again, you need to think about that. We have automation and we have standards that can help you address some of these challenges with security and compliance. So that's really strong insights, Fadzi. I mean, first of all, Ansible has a lot of market momentum. You know, Red Hat's done a really good job with that acquisition. Your point about repeatability is critical because you can't scale otherwise. And then that idea you're putting forth around control, security and compliance, it's so true. As I called it, the shared responsibility model. And there was a lot of misunderstanding in the early days of cloud. I mean, yeah, maybe AWS is going to physically secure the S3 in the bucket, but we saw so many misconfigurations early on. And so it's key to have partners that really understand this stuff and can share the experiences of other clients. So this all sounds great, AJ. You're a sharp financial background. What about the economics? Our survey data shows that security, it's at the top of the spending priority list, but budgets are stretched thin. I mean, especially when you think about the work from home pivot and all the areas that they had, the holes that they had to fill there, whether it was laptops, new security models, et cetera. So how do organizations pay for this? What's the business case look like in terms of maybe reducing infrastructure costs so I can pay it forward or there's a risk reduction angle, what can you share there? Yeah, I mean, the perspective I'd like to give here is not being multi-cloud as multi-copies of an application or data. When I think back 20 years, a lot of the work in financial services I was looking at was managing copies of data that were feeding different pipelines, different applications. Now, what we're seeing at Idaho, a lot of the work that we're doing is reducing the number of copies of that data. So that if I've got a product lifecycle management set of data, if I'm a manufacturer, I'm just gonna keep that in one location. But across my different clouds, I'm gonna have best of breed applications, developed in-house, third parties in collaboration with my supply chain, connecting securely to that single version of the truth. What I'm not gonna do is to copy that data. So a lot of what we're seeing now is that interconnectivity using applications built on Kubernetes and that decoupled from the data source that allows us to reduce those copies of data. Within that, you're gaining from the security, capability and resilience because you're not leaving yourself open to those multiple copies of data. And with that comes cost of storage and cost of compute. So what we're seeing is using multi-cloud to leverage the best of what each cloud platform has to offer. And that goes all the way to Snowflake and Heroku on cloud managed databases too. Well, and the people cost too as well. When you think about, yes, the copy creep, but then when something goes wrong, a human has to come in and figure it out. You brought up Snowflake, they get this vision of the data cloud, which is data, I think we're gonna be rethinking, AJ, data architectures in the coming decade where data stays, where it belongs, it's distributed and you're providing access. Like you said, you're separating the data from the applications, applications as we talked about with Fodzi, much more portable. So it's really the last 10 years and be different than the next 10 years, AJ. Definitely, I think the people cost collection is huge, gone are the days where you needed to have a dozen people governing, managing, applying policies to data. A lot of that repetitive work, those tasks can be in our automated. We've seen examples in insurance where we've reduced teams of 15 people working in the back office, trying to apply security controls, compliance, down to just a couple of people who are looking at the exceptions that don't fit. And that's really important because maybe two years ago, the emphasis was on regulatory compliance of data with policies such as GDPR and CCPA. Last year, very much the economic effect of reduced headcounts and enterprises running lean, looking to reduce that cost. This year, we can see that already some of the more proactive companies are looking at initiatives such as net zero emissions. How do they use data to understand how they can become more, have a better social impact and using data to drive that? And that's across all of their operations and supply chain. So those regulatory compliance issues that may have been external, we see similar patterns emerging for internal initiatives that are benefiting the environment, social impact and of course costs. Great perspectives, Jeff Hammerbocker once famously said the best minds of my generation are trying to get people to click on ads and AJ, those examples that you just gave of social good and moving things forward are really critical. And I think that's where data is going to have the biggest societal impact. Okay, guys, great conversation. Thanks so much for coming in the program. Really appreciate your time. All right, keep it right there. For more insight and conversation around creating a resilient digital business model, you're watching theCUBE.