Oracle Cloud Infrastructure - Sponsored

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

Hey, you're listening to episode, Mark. I guess we did stop counting episodes, huh? We're coming up with the big 100. Almost at 100. Yeah, so I think we're at episode 92 of the CTO Advisor. Sounds about right. Yeah. So this is a sponsored episode, and we've been wanting to have this conversation for a few months, right, Mark? At least. I think six months at least. So I've written a few blog posts on Ravello Systems, which is a solution within the Oracle Cloud Infrastructure, and then I wrote a couple of articles on Oracle Cloud Infrastructure itself.

So we have two executives from the OCI team themselves to talk a little bit, you know, deeper. We'll peel back the layers of the solution and find out, you know, the what's what of this platform. First off, we have Rahul. Rahul, can you go ahead and introduce yourself to the CTO Advisor audience? Hi there. My name is Rahul Patil. I'm Vice President of Engineering. I'm responsible for the core services in Oracle Cloud Infrastructure. And Kash. First, Keith and Mark, thank you for the opportunity.

Again, this is Kash Iftikhar. I run product and strategy for the Oracle Cloud Infrastructure. My primary job is to interface with customers, partners. Excellent. Great to have you both here. Keith, I'm just going to fire away because, you know, when I think of Oracle, I think of a database company and some other things. So for our audience out there who doesn't know, what is Oracle Cloud Infrastructure? Sure. So Oracle Cloud Infrastructure, OCI, is the Oracle's differentiated enterprise infrastructure as a service platform, which is has been custom built with the best talent, cloud talent in the world to run both scale out and scale up applications.

The benefits it provides, it provides a platform where customers can bring new sets of application and build new sets of applications, cloud native applications. They can also use this platform to bring their legacy infrastructure and sort of what we call is lift and shift or move and improve their existing legacy applications. This is a platform. This is also that is also optimized for Oracle applications, platform services, as well as our application portfolio, which is the software as a service business. So, Cash, tell us a little bit about the history of Oracle Cloud Infrastructure solution.

As long as I can remember, Oracle has talked about cloud in some type of fashion, whether it's the original concept of the net computer to the intelligent database to OCI. How long has Oracle Cloud Infrastructure been in play as it is as it pertains to your role's focus within Oracle? All right. So rewinding back a little bit and maybe expanding on it a little bit. So Oracle, as you know, has been running traditional application database for for more than four decades, close to about a decade ago.

And started from our software as a service portfolio. We decided that there was definitely a market in need to take a software as a service applications or applications that were primarily on prem in a typical, you know, sort of a customer environment and offer it in a software as a service context. So a lot of work went in to basically, in some cases, rewrite the application stack from from from from the start. And now we have a complete offering that was offered maybe like seven or eight years ago called a fusion portfolio.

We did the similar exercise with our platform as a service, which is primarily around our integration services, our middleware portfolio. And those services have been in the market and are doing really well over the past couple of years. Around close to four to five years ago, Oracle made a decision to get into the infrastructure business for a couple of reasons. One, most importantly, and first and foremost, to offer customers choices, because we already had this software as a service portfolio platform portfolio.

But people were also asking for infrastructure as a service capability. So we acquired a few companies and we embarked on our infrastructure as a service journey. Since then, we have made some some changes and additions and modifications to our infrastructure as a service portfolio. And this is where I think Rahul will add in and add more color. What we decided is that what we eventually wanted to do is build an infrastructure as a service platform from scratch, because the requirements that we were getting from our enterprise customers required us to build a platform that should serve the needs that were not being addressed by the existing cloud players at that time.

So maybe, Rahul, you can tag team and sort of expand upon that on the birth of OCI and what you guys what we have been doing since then. Yeah. So, you know, about four years ago, we architected for high scale and enterprise. What that means is we focused on the raw infrastructure. We focused on high performance networking. We've talked about non-blocking networking, which means that any two machines in a data center can run full workload, talk to each other with high bandwidth in our IAD region, for example.

Each bare metal box can do two times 25 network traffic. And if all pairs in the data center can talk to each other with full bandwidth, that's because we use non-blocking networking. So we had visions about, you know, what it truly takes for enterprise applications to move over to cloud and what it truly means for customers to trust us with their mission critical apps. And that was, you know, high performance machines, high performance networking and so on. So we started off with that.

And also, one of the visions that we had was the virtualization construct that we come up with has to work with some of the traditional appliances that we have, as well as, you know, this is what enabled us to roll an exadata in our data centers away quickly. So in 2016, we launched with bare metal compute and block volumes and object storage and so on. And then we started to move higher up the stack into some of the PaaS services and higher level DevOps services and so on.

So, Keith, tying the story together, the reason why I started off with SaaS and PaaS and then IaaS is because this is the true value proposition that we are embarking towards. Some of the components are already in place, but some of the pieces are coming together. So we started in a different track. We started from an application middleware platform services and then IaaS, which is a little bit different approach than some of the other cloud providers. What we're doing is we've now built a fundamentally scalable platform on which customers will run their applications that typically have been on-prem.

And also, our portfolio of SaaS and platform services run on the same unique infrastructure. But the other thing that I wanted to point out, which I think is critical, is this platform is completely open. We believe in open APIs so that customers can accommodate Oracle workloads, Oracle applications, but more importantly, the non-Oracle workloads as well. And I think from our perspective, we feel that we want to offer customer choices. The big question that was not addressed, which this platform addresses, is there's a lot of stuff, there's trillion dollars of investment that is still on-prem, and a differentiated architecture approach was required on the infrastructure side to basically offer customer choices to move whatever they had on-prem that would not be sort of SaaSified or on the platform there to provide an infrastructure that can accommodate those applications.

I want to clarify a point that I think I heard and just make sure. I think you said that non-Oracle database customers will still have a lot of benefit from using this stuff. It's not just strictly for Oracle traditional customers. Is that what I heard? Yeah, that is correct. I mean, if you heard what Rahul said, right, we have optimized the infrastructure. In fact, there is no other infrastructure that can accommodate our converged appliances, highly optimized leading databases such as the Exadata.

And we recently launched our autonomous database strategy. All of that can only be implemented, run, and optimized on the Oracle Cloud infrastructure. But at the same time, customers can bring their own database of choice and run it on our bare metal or different VM shapes, right? So there's nothing prohibiting them from running those applications and those databases on our platform. In fact, some of these open source portfolio are also offered as a service natively from our middleware team. So think of this as an infrastructure that's optimized to run Oracle workloads for sure.

But customers can also run their operating systems and their applications of choice and their database of choice. And because we want this to be a platform where customers don't feel locked in. And that's a very, very important design criteria that Rahul and I can expand on. Yeah. So there's there's something I want to pick apart because I can see the value of being able to lift and shift. You guys use that term lift and shift my application that really doesn't fit traditional cloud whale whale, whether it's one of the other major cloud providers or some off brand cloud provider.

There's there's just some applications that don't fit well in the cloud. So the ability to lift and shift that to bare metal or to virtualize infrastructure, that looks a lot more like what I'm doing on prem makes sense from that perspective. But I watched and I participated in cloud fielding. I think it was three or four and some of the pushback that you guys got. Well, how is this not any different than hosted than a host that not cloud provider, but hosted bare metal service provider?

What what is the value add on top of just any other colo solution? If you look at our services. So I'll give you an example. So today, a customer. We're not just like a hosted metal service. Like, let's take some examples. You want to bring your database in. You can start with a single core VM and you can go up to bare metal service and you can go off into your systems. So if you think about your journey, as you know, we're on some tests with VMs and block storage and then you start to take you like you start to like it, you start to play with more systems and we give you a tremendous amount of flexibility and compute, not just in compute, but also storage.

So, for example, you can do four hundred thousand IOPs on block storage in with a single bare metal. Now, if you think about our blog, we're going to interject there for a second, because you said something that's really that I don't want to go past. And I want to make sure that we capture this because this would be a critical difference. Let's say I go in with one core VM. What is the effort to take that one core VM that's using block storage on the back end?

And I say, oh, wow, this is perfect. Use case for the cloud is a development idea. I threw out some small resources to it. Now I need to I need to go up to 12 cores or physical machine and faster box stores with four hundred thousand IOPs. What is that? What does that transition look like? So there is some engineering work as they transition from one shape to another shape. But there are things that become pretty simple. Like they're all backed by the same block storage device.

The block storage devices have an independent life cycle to the compute machines themselves. And especially, for example, if you go from one one core VM to a bare metal server, you can detach the block storage from the one core VM and attach to a bare metal compute box. Now, you have to do some work on the database administration side to be able to recognize these block devices and set them all right. There is some engineering work. And over time, we're going to make it a lot more friendly to be able to, you know, quickly click and be able to scale up and scale out.

Right now, there is a little bit of engineering work, but it's not too bad. Our customers, you know, especially they love the fact that they're able to test in the same environment, their POCs and network environment the same. The APIs are the same. The underlying devices infrastructure is the same. And they love the fact that they can scale up and scale out in the same environment. Yeah. And let me sort of expand on that. I think this is a really good question.

So what I will talk about was a customer that started on a particular VM shape, and now they want to transition over to a bare metal or vice versa. Right. Typically, in the customer engagements, what we are finding is as we as we offer customers to run their applications, either cloud native or the existing application and sort of move and improve them. That's a new word for lift and shift. Believe it or not. Over to our platforms, they generally have a pretty good idea in terms of specific workloads, what they would use.

Generally, what we are seeing is for databases, as an example, they would probably go for a bare metal offering and they will keep on scaling the core processors as needed. And then for some of the the application of the middle tier apps, they will probably go with some specific VM shapes. But from our perspective, I think when we think of compute, we think of compute as a way for us to offer different choices that customers would use. You know, they can use bare metal, they can use VMs, they can use containers and very soon functions.

Right. So from our perspective, it's a complete cloud offering like any other cloud providers. Yes. In some cases, when they're moving from like a VM to a bare metal, as Rahul mentioned, there's some work needed. But from our perspective, customers have different choices and they have a pretty good idea of what workload, which will run better on what shape or bare metal or a container or run it as a function. Now, I'll give you one example, like on the engineering side, like how I mean, this is truly differentiated in all the competitors has and how we are reducing the downtime for, you know, cloning and creating installs of those clones.

So, for example, on block storage, we offer a coordinated snapshot and restore, which means that you can define a block storage group and say, hey, this compute instance, this boot volume, along with these four block storage volumes attached to this compute computer. And also you can also define a cluster of computers and say, OK, all of these devices together, take a point in time snapshot and then clone that environment. So like, you know, we have the most differentiated offering from a block storage perspective where you can coordinate across multiple block devices attached to multiple compute.

So you can do a coordinated clone that makes it easy for you to replicate or move your environment from one thing to another. I just want to give you an example of engineering what we're doing. So you mentioned that a lot of your customers already understand where their workload may best fit. And that kind of made me wonder, because in my experience, customers don't always know, you know, should we lift and shift, should we refactor, should we go cloud data, should we do whatever?

So that made me wonder, who are your typical customers today? Who are your target customers, actually? So I can give the answer to you in two different ways. One is that typically if you look at the Oracle customer base in general or the customers that we have, you know, we are fortunate to have customers that are already running Oracle applications, so very high and, you know, Fortune 500 customers worldwide, right? That's where we spend a lot of our time. Then we have our customers, which are sort of enterprise customers, but not as much scale as the large enterprise customers.

And then a place where we're seeing a lot of traction, in fact, where we're seeing a lot of new customers come in is what we call our small-medium business customers, right? These customers span across different verticals, so retail, financials, you know, healthcare. And as you can see, Oracle has been selling databases and applications to these customers for a very long time. Where I think the discussion that has been happening around primarily the high-end bare-metal service, for example, where people use high-performance computing, you know, it's across all the segments and across all the different verticals.

For example, for bare-metal in particular, people are using that for high-performance compute applications. In cases where they're thinking about, you know, coming up with newer applications or building newer applications, we generally see those conversations mostly happening primarily in our SMB space because typically the customers are smaller size and they want to experiment with some, you know, applications that they want to just kickstart. But on the high-end of the customer and the enterprise customers, what we're seeing is that, you know, it's a combination of both some existing applications.

Typically, we have the lift and shift or move and improve motion, where they have an existing database application that needs to move to that cannot be refactored because these applications are quite complex, tied to an Oracle database where we offer these migration services. But then also they have a departmental need to run some, you know, sort of cloud-native applications that come to our platform. So I think it's weighted across different verticals. I think the common theme over there is that they're realizing that this offering that we have, they can start with either, you know, bringing a new application or they can either start with an existing application and then expand from there.

Yeah, we definitely have a customer base where we start with a, you know, sort of a move and improve sort of a customer use case. But I just wanted to clearly call it out. There are two or three specific use cases that we are seeing. One is definitely in the move and improve where there's an Oracle application tied to an Oracle database or a third-party application tied to an Oracle database. And this is very important to understand, and it leads to answering your questions around, we need to provide interconnectivity to different systems because we have a huge ecosystem of partners that are running their own applications.

These are not Oracle applications. In some cases, leveraging the Oracle databases, right? So that's the second category. And the third category on a broad level is a set of customers who are just building brand-new applications, right? These are, you know, typically in some cases they're using, you know, container services that we just launched. But also what is differentiated is we are seeing a lot of high-performance compute environments running on our platform. In fact, by Oracle OpenWorld, we'll be publicly mentioning that one.

But one of the largest customers that is the largest customer, in fact, running on our platform, is actually a high-performance use case, right? Because they started slow and then expanded upon really, really, really, really fast. So I'm going to touch a topic that with some cloud providers might be taboo, but I have a feeling that with you guys this won't be the case. We live in a hybrid multi-cloud world. And there are simply, you know, you guys have your focus. Your focus is on traditional enterprises who cannot easily take applications and migrate them to the cloud.

You have a land and expand or land and improve, move and improve strategy to help those customers. However, there's just services that are best suited, whether it's for geography reasons or service reasons and other cloud providers. How do you guys work with or integrate with other cloud providers? Now, what we're finding is as we service these couple of use cases, in some cases there is integration services of these applications with some other services that may be running in other cloud providers. So what we do is, of course, within our fabric, we have, you know, a backbone.

We have fast connectivity or fast connect services that basically provide connectivity from our regional POPs or our, you know, regions to the customer specific sites. And in many cases, people can leverage the same technology to connect to other cloud providers. So what we are doing is we are providing open API connectivity to our platform. We just enable them through, you know, some connectivity option from a networking perspective. And they're able to land on a platform. And this is important to understand because we are also seeing now examples of customers transitioning their workloads or interacting with their workloads.

And specifically as an example with Amazon. And we are seeing them basically take the VCN, our version of VCN and a VPC and connect it together and run some of the workloads on Oracle platform. And, in fact, migrate some of the data from Amazon over to us and vice versa. So I think this is something that you would expect from us. This is something that we are providing. Rahul, you want to add to that? Yeah. There are multiple dimensions to, like, you know, how do you make hybrid cloud work successfully.

Okay. So one is obviously connectivity. And I think Cash captured a lot of the connectivity options that we have. I mean, you know, primarily it's a FastConnect, but it's the choices that we have in FastConnect that customers already are there both to connect, I mean, to connect from on-prem to the cloud. But when you talk about hybrid, developing for a hybrid cloud or multi-cloud strategy, I think a lot of our customers, this is where Oracle, I think, is doing, you know, we picked the right strategy going forward is we embraced open source wholesale.

Like, you know, the fact that there was, over the last few years, there have been clear winners and clear leaders in some of the open source technologies when it comes to deployment, orchestration, streaming service, and things that, you know, help get data from one side to the other side and all that. And we're embracing in all of these cases, we're embracing the open source strategy, which is, you know, Terraform, use Kafka, you know, those are the things that we are embracing. And it's resonating a lot with customers because, hey, they have to manage, they need all these tools to manage on-premise, and they've already embraced all of these things.

And, you know, when they start thinking about multi-cloud, they love the fact that we made it. Most customers have embraced it. We'll be sending our notifications through the Kafka streams and so on. And so they really like it, and it really helps them target multi-cloud in a very friendly way. With that, are there any customer stories before we close out that you guys wanted to highlight before we end the conversation? And I can pick up one each from different categories, the one that I mentioned.

So I can start off with Darling Ingredients. It's a sustainable ingredients industry. And when we engaged with them last year, and they're a public reference, so I can mention their name, they had the problem that their requirement was actually, hey, we want to get out of our data center. We have an investment in Oracle databases. We have an investment in some application that we have built that leverages the Oracle databases. And we use some of the Oracle technology as well.

I think in their case, it was like EBS. So there was a move of consolidation. So consolidation of databases at the same time they wanted to shut down the data center. So they consolidated the data spaces from 19 to 3. We used technologies that Rahul mentioned, you know, FastConnect. We took their services, their databases. We moved them over to our cloud. We did the migration of their applications over to our platform. Then what we did is we also helped them and now helping them build their new applications on our platform.

And that transition took a couple of weeks. And now they're happy and live in production and our reference customer. That would be one example. And the segment specifically for them was they're in the enterprise segment. The other example I would give you is of actually a startup. Because sometimes when you think about Oracle, you generally think about large enterprise customers. And you think about use cases you were mentioning from earlier on, you know, move and improve or lift and shift.

But YellowDock, we mentioned before, was a customer that we acquired very early on last year. And the reason why I bring up their name is that this is an example of a customer that's a startup, a pure startup, based out of Bristol, UK, and started off using a couple of VMs and a combination of VMs and bare metal servers. And what they do is they are building a platform on OCI. And what we had to do is we learned a lot through that exercise.

We actually, you know, did a lot of modification on our pricing model, which we haven't talked about, but I can. To offer startups, you know, a comfort that they can come use these really, really high-end services at a minimal cost point and then scale. And right now they are using thousands of cores on our platform and, in fact, servicing their end customers. Another example that I can give you is of OceanX. This is in the digital, you know, marketing space. And this is an example of a customer that was running a complete non-Oracle stack end-to-end on a very large cloud provider name, let it be unknown.

So this was a migration of a service that was actually running on a competitive cloud platform. And what basically got them attracted about Oracle is that they were using a database service on the other cloud. And what we did is we took a snapshot of their dataset, moved it over to our Exadata platform that was running on OCI. And we ran some tests for them, and they saw about 10x improvement in their queries on the databases that were running on the OCI platform.

And then we basically used our migration techniques to basically take their workloads and move them over to our platform. So right now, Keith and Mark, we have our customer count is increasing exponentially. So I just gave you an example of three. If you look at Oracle Open World sessions from last year, and they are publicly available, we had more references. We have about like close to about 30 to 40 customers on stage. And this year, we are going to do the same.

The last example, sort of closing it up, I want to give to you is in two segments, which I think are very, very important. And I want to make sure that we call it out. We are spending a lot of time in our partner ecosystem. And what we are finding is that the partner community, one, you know, who have been traditionally engaging with Oracle, in some cases, these are the net new partners. What they're doing from their business perspective, and the partners are in the, what I may call it, the ISV, independent software vendor category.

And the second one is the system integrated category. Both of them are going through an interesting transformation where they themselves used to have either a consulting process on-prem and they used to provide hosting services. This is when I'm talking about SIs. And now their business is sort of transforming into people who want to move to the cloud and they want to still offer sort of managed service provider offerings on our platform. We are now seeing a significant sort of ramp of those customer engagements through these partners who basically have invested time and energy to build their service offerings on the Oracle Cloud Platform.

So we have Guest, SmartBlog, a couple others. And from a global GSI perspective, this is publicly mentioned. We have a relationship with Accenture, Deloitte, Cognizant, all the others. So we are seeing through them enabling, enablement through them. We are addressing a very, very large customer base. And that work is going really well. com. The last segment and probably one of the most critical segments that we are finding, which is spurring the growth of consumption on our platform, is the independent software vendors.

And what was happening is that, you know, you had people building applications tied to either Oracle databases or not. Where Oracle had a relationship and these companies are servicing different verticals like retail, healthcare, stuff like that. And what they were doing is they had an on-prem offering. And now their customers want to either move those applications over to the cloud. So vis-a-vis classically, you know, our move and improve story. But then they are also creating these multi-tenant applications or SaaS-ifying their application.

A typical example would be Manhattan Associates. Their application is used in warehouse management systems worldwide. So that's a typical Oracle partner. We are getting or we are servicing our end customers through enablement of these partners. Another example would be FireEye. FireEye publicly announced their email security system now running on our platform. And we're working very actively to get more of these publicly available references available because we have a significant footprint of these customers and partners and ISVs now. That we plan to roll out and be able to publicly discuss with you in the next coming months.

You know, kind of maybe hard question. What type of adoption are you seeing from your existing customers? So I think the existing customers, the adoption is very promising. From last year to this year, because as Rahul mentioned, the first time we launched our first region was in the September timeframe of 2000. I think that was 2016. And truly it was an MVP with most of the services that were enabled on the OCR platform was primarily in the June timeframe. Correct me if I'm wrong, Rahul.

I think June last year was when we truly had like two regions, at least in the United States. And we had most of our features available on our platform, including our platform services. So the services have been available for close to like a region, close to two years, but truly like about a year and a half. In this timeframe, we're seeing exponential growth. And like I'm saying, like I was mentioning, you know, for majority of these customers, which are Oracle customers, these are existing applications that are moving over.

But when they land, what we are finding is that the perception problem, some of the, what I think Keith, you were mentioning, the perception problem, hey, it's Oracle. It's just built for Oracle applications. I think that's fading away. So what we're seeing is what has, we are getting more and more of these customers. We're expanding into new use cases, which they initially thought would not be possible on the Oracle infrastructure. Because the most important thing that they are trying to tap onto is data.

And a lot of this data runs in Oracle databases. And all the applications are basically servicing and getting inputs from this data. So that's sort of our competitive advantage, but we have a lot of work to do. And we have to execute, but all signs are really positive. Well, Raul, we really appreciate you guys coming on and helping us better understand the OCI platform, its position in the market, and basically the go-to-market strategy. If folks want to find out more about the platform or follow either you on social media, how do they do that?

com. We are also very active on social media channels, both on Twitter. So it's Oracle IaaS is the channel that we use, or you can even reach out to us at Oracle channel on Twitter. And then we also have presence on LinkedIn. I'm also personally available on Twitter. I'm very active on LinkedIn. So feel free to reach out to me directly, Raul. I'm on LinkedIn as well. Keith, you can find me on Twitter. It's at SensiStorage, LinkedIn, all kinds of places.

All right. And you can find me on the web at CTO Advisor on Twitter. com. Talk to you next. CTO Advisor podcast.