Cloud Crunch
Cloud Crunch

Episode · 2 years ago

S1E13: Examining the Cloud Center of Excellence

ABOUT THIS EPISODE

What is a Cloud Center of Excellence (CCOE), and how can you ensure its success? Joe Kinsella, CTO of CloudHealth, talks with us today about the importance of a CCOE, the steps to cloud maturity, and how to move through the cloud maturity journey.

Involve, solve evolved. Welcome to cloud crunch, the podcast for any large enterprise planning on moving to, or is in the midst of moving to, the cloud, hosted by the cloud computing experts from Second Watch, Ian will be cheap architect cloud solutions, and Skip Berry, executive director of cloud enablement and now here, are your hosts of cloud crunch. Welcome back to another episode of cloud crunch. I am joined today by Rob Wheelan as my cohost today, as skip berry is taken so much needed time off. Today we have a very exciting guest today, Joe can sell, a VP and CTO at cloud health, which is a vmwhere company. Joe's and entrepreneur and technologies who's passionate about Boston and Cape Cod, I suspect, baseball and all things cloud, not necessarily in that order either. He found a cloud health in two thousand and twelve with the goal of disrupting the growing complexity of cloud computing. The company was bought by vm where in October two thousand and eighteen, after becoming a global leader and public and private cloud management for enterprises, partners and SMB's. Joe was previously VP and engineering at Amazon, backed cloud archiving company, Sony it, where he was a pioneer of using multi public clouds at scale and also building one of the earliest cloud centers of excellence, which is what we're to be discussing today. He was also a managing director at Dell, where he led global engineering teams delivering multiple software as a solution products or SASS products, and VP of engineering at silverback technologies, which was acquired by Dell, where he helped pioneer remote iteam management software. Joe also has the unusual distinction having been a member of the first scrum team as a founding member of the EASEL SICRONICITY team before the advent of the Agile Manifesto. Joe, welcome to the show. Well, thank you and glad to be here. Awesome, awesome. Yeah, so today's topic is going to be a...

...cloud centers of excellence and all things related around out of cloud, and again, thank you so much for joining us today. Excellent Joe. Tell us what is a cloud center of excellence? That's a great question. I think the industry has put a lot more definition, I think, over the last several years around a cloud center of excellence. It's usually called the clouds iner excellent, but I've seen many different terms for it. So I've heard many companies also use a cod business office. In some cases organizations just call it their platform teams, but basically it's a group of cross functional leaders who get together and together they start to drive the best practices and standards that govern the cloud implementation across their organizations. And this is really driven from a change that a lot of people don't fully appreciate that has happened in the cloud, which is precloud, all all of our usage of infrastructure and all of our deployments of applications were all controlled by central it. So your it organization made available the infrastructure and or the applications and then had control over managing them. And in the post cloud world, management in a large enterprises occurring in some cases in a thousand or more places across the organization, as opposed to being located solely in central I. And I think that change is really a seismic shift and it's it's brought both agility and innovation across organizations, but is also brought risk, and so cloud center of excellence is a way to manage that risk. Awesome, that's great. How Could Cup is it sure that they had the elements in place for it to be successful? Yeah, I think it really starts with first and foremost, with the recognition that you actually need to build a cloud center of excellence. So you'd mentioned in my intro. You've mentioned that I'd built one of the early cloud centers of excellence and that was in the two thousand and ten eleven time frame. And what had happened in my experiences I had really scaled up my company's use of the public cloud. In fact we were one of Amazon's largest customers back in the two thousand ten...

...eleven time frame. And then, after scaling it up, I ended up bringing a multi cloud across multiple other public clouds and what I found is that the cloud just moves at a pace that is very different from the past that occurred inside traditional data centers invected. It occurs at a pace that in many cases is forty five orders of magnitude faster. So if you step back, you think of one order of magnitude, you know, if I add another zero on your bank account, that that seems like a pretty size with change to you, but you had four to five Zeros on top of that. That bank account and that's a big difference. And so what we saw is that the cloud moves so quickly that you need a way to govern. So I used to in the early days of building this out, I used to call it distributed management, centralized governments, and so the goal is to bring a group across functional people together from finance and operations and architecture and security and to allow them to not dictate, not control, but to put in play standards and best practices and make all the places where management of the clouds occurring more effective. It's great. So now you did this, and what year? Going back, I joined Sony and as VP of engineering in two thousand and ten. We went multicloud in two thousand and eleven and over the course of series of several years ended up on or different clouds. So and cloud years. That was about seventy seven years ago. By the math correctly, that's exactly right. So looking at that, obviously you know the clouds changed. Right, it has, and it has it there's been a lot of new capabilities. Do you think that model what we did then? Would you change anything different from then to now? Without a doubt, I think at the time I think it. You really have to adapt your cloud center excellence to the needs of your business and you also have to adapt it to the state of your maturity, and it's I think there's a long conversation we could have if we had more time, around just maturity. But in our organization we were a tech company, so we had different needs as a tech company. We were going to market through a channel, so that kind of drove a different business model. That really impacted how we thought about parnessing the power...

...of the cloud and we had a very clear strategy of how we were using the public cloud and how we're using multicloud. I think in some organizations they have greater scale, they sometimes have less strategy, they have different business needs. So I really think you need to build your cloud center excellence around first that recognition I talked about, but you need to get the organizational commitment that over time your needs will change around what you need from a cloud center of excellence and what you expect from your cloud center of Excellens. So I think you have to adapt it to the specific organization, but I think it starts with, you know, the recognition you need to do this to scale, and I say this to all companies that I meet where I'm meeting a leader who's scaling in the public cloud, and I will tell them you will hit a wall if you do not put in place best practice, the standards, if you do not invest in the cloud center excellence, there will be a tipping point at which that easy agility, in the easy innovation that you received by leveraging the public cloud will suddenly turn against you and so this becomes a mechanism that it's not discretionary, it's actually a prerequisite to scaling in the cloud. So you mentioned, Joe, that you need a cloud center of excellence because you're managing a large number of people. Also, the pace of the cloud, it's just fast. Things keep changing. Is there anything that's a constant thing saying and governance, anything that it would be a constant from two thousand and ten to now in terms of what people should be looking at? Yeah, it's interesting you say that. So there's really three areas that I focused on in two thousand and ten and eleven and those three still today represent the foundation of a cloud center excellent. So the first is financial management, which is, as you go to a public cloud and you start to try to unlock the innovation and agility of a public cloud. It comes with a risk of budget overruns and you know, in fact, you see many cfos who bought into the strategy of the public cloud only to find that they're paying for data centers and now they're increasingly paying for a really large and growing pub...

...of cloud bills as well. So you really need to be deliberate in terms of how you do financial management in your cloud center of excellence, and that's something that you can't do across a hundred or a thousand different places across the organization. It needs to be centrally governed seconds operations, which is really to do really harness the power and scale in the cloud. I think you need to put some standards in best practices in place. Often these start around configuration and it's around it's around tagging policies, it's around reference architectures of for these workloads. We spin up these types of virtual machines with this type of attached storage and these types of performance characteristics. But standardization becomes a prerequisite to repeatability and really drives your ability to gain the best Roi from the cloud. And then the last time would just say a security compliance. It's an area where, when you run a cloud speed you know, you really need to know that you have confidence and integrity, that all those places where the cloud is being managed day to day and changes are being driven day to day don't end up putting you on the front page of the Wall Street Journal. Yes, that's it's often not where you want to be, not for the wrong reasons, what I think. So I do want to make sure our listeners really understand to is that you talked about being a tech company and how involved you brought in finance, and I think this has been a successful model. We've seen it over and over again as well, and I just don't think we can read it. Are It enough? I think a lot of people look at and say it's a painful thing to do, but I know then you have this experience to you can get them there where they're all contributing equally, all those stakeholders. That's right. I mean there's been a revolution in finance, I think, over the last five years in terms of finances engagement in management of cloud and infrastructure, and initially we didn't really understand how finance could effectively engage, but I think now there's an emerging model guided by the Cloud Center of excellence that I think is becoming highly effective. In you know, we're seeing organizations go from leveraging the cloud and their justification is agility innovation,...

...but it's coming at a cost to increasingly organization is actually getting better economics and being able to do things in the public cloud they can't do inside their data centers. Yeah, that's great. So a couple recurring things I see in your background is actual and I see cloud and in today's environment, you know, with the CCOEAR Class Center of excellence, how is that still so important? Yeah, I think if you look at my background, you see at one point in time I work for Dal through an acquisition of a company, and I remember at Dell we used to have this annual cycle we would go through of planning our infrastructure for the next year. So We'd have our business plan, we'd have, you know, what we want to do achieve from a roadmap perspective and we go about actually deploying in the data centers what we needed and it would take some cases weeks and months to actually procure the equipment and then rack and stack it and then put the operating systems and the applications and and there was this rigidity to the process. But that's how things were done a decade plus ago. Right. I think what happened with the cloud is the cloud was not something that God adopted by central it it would. It got adopted by a team in a department in a line of business and then from their started to spread almost virally across an organization. And I think one of the things that we need to appreciate is that that freedom to go drive those decisions, that freedom to make the day to day management decisions, the freedom to actually provision what you want when you want it, is part of the agility that we get from the cloud. So so the key of a cloud center of excellence is not to over rotate, which is that freedom. Is What is giving you the innovation. It's one of the powers that comes with the cloud. You have to strike this balance of providing the agility and innovation and all the benefits of the cloud, but doing it a way that is is not adversely impacting your teams in their ability to get things done for the organization. So this shift from Hey, I'd like to lear of the cloud for innovation and agility to finally maybe getting some economics benefits out of...

...it. What are some some common trends are saying there? How didn't do that? What company is and what teams are actually getting those economic benefits from the cloud? Yeah, I think the places where we find that you don't get the economic benefits from the cloud tend to be traditional application. So if you look, we've we've kind of lived for the last couple decades with an architectural model in which the operating system is kind of central to the way that we actually build and deploy and manage applications. And when you look beyond the operating system, whether that's containers or just the rich array of platform services that you see available today, you start to see opportunities to actually leverage the cloud and ways that you couldn't actually leverage inside your data center. And so so, as an example, at my previous company we really leverage the spot market place. So we had some workloads that were heavily analytic driven and they needed burst capacity, and so we used to we would automate acquiring the optimum price for compute and then discarding it, and so we basically built by the boat and workloads where everything was disposable and everything could follow peaks and valleys in terms of demand. And so when you start to do that and you start to align supply and demand in a way where it's reactive and responsive, I think you start to really unlock the power of the cloud, because you're not making the capital expenditure for the infrastructure that you you have available to you. You're just consuming it only when you need it. And so I think today, when you think about this, ten years ago, when I was doing this, it didn't apply to every workload and most workloads, most open source, are built around the operating system. Is that, you know, central component of which of your application architecture. But today, increasingly there's a lot of workloads where you can take advantage of that, and the more you start to actually build cloud native or cloud centric architectures, I think that the more potential...

...you have to actually unlock the financial potential. Yeah, I think what I've seen in cloud consulting for the past s years is people who come straight into the cloud with sort of legacy workload are going to be disappointed. Usually in economics you just run the way around in the data center. It's just going to be costly and you're not going to have those awesome cloud benefits. But interesting is if you really sink for that agility, so for example, of containers or or Servilis, that's where you actually get the economic benefits. Of course it's sometimes costly to get there if there's a lot of refactoring that sort of thing. Yeah, that's exactly right. It's my experiences as well and I think sometimes people they believe that they're going to the cloud to actually save money and that's actually a journey. It's you know, it's hard to achieve that destination. It takes a lot of investment and it takes a lot of rearchitecture or implementing new architectures to be able to actually achieve that. So I think your spot on. I think that's been my experiences over the last decade as well. So can you tell us more about cloud of maturity? So maybe there's four steps of cloud the charity tell us about that? So we cloud health. We've put forward this coug maturity model and it's interesting because it's based on Tenzero customers that we've worked with in the cloud. It's based on patterns that we've seen across these customers and what's worked for them and in the maturity model, like a traditional maturity model, the bottom left is all about being immature in the cloud, in the upper rights being highly mature in the cloud. But what we've seen is is there's effectively three areas of investment you need to make the scale and mature in the cloud, and those are what I mentioned before, which is it's financial management, operations and security. But there's there's four phases that you go through in each of those areas of investment and it starts at the most basic level. It's starts with visibility, and visibility is a is often about what am I running? Why am I running it? You know, how much is it costing me? How does it...

...aligned from a business perspective? It's you know, visibility gives you the connective tissue that allows you to make smart decisions, but it doesn't actually make those decisions obvious to you. The next phase after visibilities all around optimization, and so this is, whether you know, in the financial management side, it's around how can you make a workload be sized appropriately so that it's not overcapacity but yet can support the demand that it has? In the case of security, optimization can be how is it that I proactively monitor across all the changes that are occurring today, which could be hundreds of thousands, millions of changes across an enterprise, and make sure that the changes all fit into our best practices, in our standards, in our policies? And then, after you get to optimization, we find that companies really struggle with scaling up this model in the cloud. And so the next phase we call governance in automation, and this is around moving away from just pushing out more dashboards or notification alerts or reports to your stakeholders and it's about how do I proactively monitor for the ideal state I want my workloads in my applications, in my business services in and then how is it that I automate the outcomes that I want? And so this is about removing the people, removing the human error from it and trying to actually keep in the optimum state all the time, or nearly all the time. And then the ultimate state that we really help steer our customers towards is what we call business integration, and this is where the cloud gets integrated into your interpret Dash Boards, integrated into your service catalog, and it's it's really connecting out across the organization. It's no longer the destination of the cloud that you're focused on. It's just part of how you transact your business. So shifting gears a little bit. There's been obviously cloud certam excellence is key to all those as well, but there are other models that people have started to talk about. They may or may not cover all the same things. One of them as the cloud enable a engine. We're seeing some more...

...nomenclature around that. Where some of your thoughts on that? I mean, I think there's a number of different models. Some of them are are focused more around how you drive an organizational transformation, some of them are more around governance and scale. I think I wouldn't stand here to tell you that there's only one model that applies to all of your problems. I think where the cloud center excellence fits best is as you're scaling your usage of the cloud across multiple teams, especially if there's any organizational complexity. I think that's where you really invest in the the cloud center of excellence. It's a way to actually successfully scale without putting your business at risk, without having cost overruns, without impacting SLA. That's the place that I think the CC OI fits, but there's other models, like you know, you look, I think some organizations are trying to drive an organizational transformation around the cloud. That's not what a CECO he's all about, right, you know. So there's other models that apply for different problems. Fantastic. Okay, given all that Joe, I mean I go to the turn of I think was really interesting, especially the business integration. That's kind of like cloud nervot it. A lot of people on vision, but I don't think you can skip over any of those stayses. Given interview of the last got winner is excellent. SCURE organizations focus on to sort of measure club maturity. I think what I would do is I think there's different metrics that we recommend and see make organizations successful and each of those different areas of investment. So I think there's metrics that we recommend around financial management, around operations as well as security. But there's also metrics you can look for, whether it's visibility optimization, governance and automation or Business Integration. But I think they change as you mature. So, as an example, a metric that you might use for a financial management when you're really at the visibility stage might be as simple as how much am I spending in the cloud and how does it aligned by team or cost center or Department or Business Unit? Whatever. The perspective is that you actually want to look at that...

...data buy, but by the time you actually get the business integration it's more around integration with your KPIS. It might be what's my cost per document per month or, you know, what's my cost per user per month and how is that changing today or this week? So I think there's a set of metrics you can use, but they alter as you start to mature. Yeah, I can say that that kind of height integration being absolutely amazing. Different stories of customers is seeing to that really well, I have you know. And so I actually when the story of Sony and actually did this myself, and it was kind of an interesting story, which was when I came in. I was brought in as a turnaround vp of engineering in this really fast growing start up that was entirely based in the cloud and we reached a point where the cloudspend was going up rapidly and I had a month where I got a three hundred and fifty thousand dollar at Os bill not long after joining, which I can tell you in two thousand and ten, early two thousand and eleven, was really substantial. You didn't have a lot of companies doing that today. Less interesting, right and and I did the analysis and looked at it from a business perspective in the fundamental metric by which we were actually driving the business was cost per document a month, and what I was finding was the cost per document per month was running around two dollars and four cents per document per month and I thought, well, that's interesting. Date and I went back and I did the analysis and found that it was around eighty four cents. Was the break even point from a business model perspective. And so effectively we were running a business that was just going to lose money in the COT if there wasn't a substantial change in direction. And by the time I left I had it down to around thirty two, thirty four cents per document per month. And it was this, you know, seismic transformation. It was an example of a business model that was not going to work in the cloud to a business model that actually had a competitive advantage by being in the cloud. And so I think I've seen that story myself first hand and I've seen examples of other customers who've been able to drive that type of business transformation.

That kind of story is great. That's a huge decrease in costs and that's near and are of our art. So I wanted to ask a question to tie that back to, you know, the original definition of the cloud center of excellence financial operations and security. So culturally, how do you how do you get that story of like Hey, here's the end goal of business integration? How do you get people who are either all the clause Sarah Excellence or will be on the clouds center of excellence to embrace that that end state? Do you find that they're they adopt that easily, or does it take a little bit of work? Yeah, it depends on the organization. What we have found in general is at the most basic level, when you're at the visibility phase, it's just you provide the information. So you bring the information. Often at a phase of visibility you're bringing it down to key stakeholders across the department or a team, but not down to say, individual developers who might be on a team. But your goal should be, you know, initially to start to push the information out, but not necessarily with the point of view you know, not with this is good, this is bad. It's just a start to a raise. In the case of financial management, I call it the the cloud cost Iq, which is, you know, to the extent that you can start to raise that cloud cost Iq and that awareness of costs, that's really the beginning of it and then over time, I think the most successful organizations I've found is you need to have a bit of Gamification of where people are getting regular report cards and they're actually seeing their peers and they're seeing how they're scoring relative to their peers to be able to really start to drive change across an organization. But I think it really requires a commitment top down to actually commit to driving a cloud center of excellence and having best practices and standards. Without that it's often hard to get the ground up support. You'll get some level or what we've seen as organizations that...

...don't have the top down commitment, they'll have pockets of excellence across the organization and then pockets where it's not occurring. And I think in the case of costs, that could take a public company from, you know, having a good quarter to having a bad quarter. If you do that, it could take a company that might be in a high, highly regulated industry from from having a security breach or not having a security breach. So having not uniform adoption, I think, is a great risk across the organization. So I do think it starts with executive level comitment to building and investing at the CECOE. So looking forward, I don't believe there's really an end state in this transformational time period. Things keep maturing, there's new features, there's news technologies. What do you think's next? Yeah, I think we do live in the industry that's sort of the gift that keeps on giving, which is a constantly changes, and I guess the only thing that we can rely on is it's it'll be different tomorrow than it is today. I think the things that I've been paying attention to is in terms of trends that I think are real big drivers of the next few years. Is certainly platform as a service. So when I started the company, Ninety eight percent of what was occurring in the public cloud was actually effectively virtualization. It was virtual machines and virtualized storage, and that really was ninety eight percent what we called the public cloud. And then the other two percent was mostly as three and you look today like that's completely shifted. And I'm not saying that the dominant paradigm is actually platform as a service, but I'm beginning to see this change in terms of not just adoption of platform as a service but also starting to treat it as almost a set of colors by which you can actually paint a Pallette by which you can pay. And that there's you know, I'm seeing increasingly seeing organizations look at it not by vendor, you know, it ws platform services, UNI versus Google, versus as your platform services, but all of these comprising a set of underlying building blox upon which they can bring applications to market. So...

I've seen companies that that actually combine as your Google, a Tobos all into a single business service and all would delivered reasons of why they've chosen those platform services. So I think that's an exciting trend, which is to go beyond the barriers of a public cloud provider but also start to leverage a rich platform services where the operating system isn't the core of our application on architecture anymore. And I think from there, I think what you'll see is is as that emerges, I think two trends will probably occur. One is a trend I call microsass which is, I think increasingly we're going to see third parties come out with very verticalized solutions that build on top of these platform services that solve specific problems of you know, whether it's a marketing or customer relationship management for us, that we can use to build upon. But I also see a trend of I think we'll start to see more third party platform services that will be competitive, like a better solution to Dynamo DB or, you know, better really vertically focus machine learning services. And so I think these are all trends that I'm watching and I think this change is accelerating to actually leveraging the cloud. In the analogy I used to give years ago, what I would do public speaking was I used to use the analogy of the Digital Camera, which is for anyone old enough to have seen a digital camera, which is before Digital Camera, you had film based cameras and I remember the very first digital camera I use. It had the exact form and funk suction of a film based camera, but it was digital, like the buttons were in the same place and everything and this is a classic trend that occurs with version of it tends to look like the predecessor technology, just like the cloud look like virtualization when we first saw it. But the two point on innovation is where the real disruption occurs. I think we're just getting started with a real disruption. Great, wow, that's that's really interesting. I mean, can you expand on this? The Color Palette Concept? Is that when you see as part of this to point out? Yeah, I do. I think a lot of people are struggling with, you know, vendor lock, in no vendor lock, and I see world tomorrow where where services are veiled from many...

...vendors and it's not like the big three public cloud providers. You could have, you know, small startups that provide one, two, three, critical services to to the market. And so I think, you know, I see this world where we're just going to assemble those different services and I a couple years ago I was at a cocktail hour in San Francisco. I was up on a, you know, rooftop and it was a medical and imaging application and it leverage Google machine learning, with s three, four, storage of the images, with a set of other third party services, all brought together to actually deliver medical imaging and it probably had five or six different vendors in the stack, all using different platform services. And I think what enables this that wasn't available to us today was the impediment. It really was always network costs because you needed to have your compute in your data close together. But I think is we look for microservice applications are really starting to break that barrier for us, which is now data is smaller and data can be in, you know, smaller units in closer proximity too, lots of different areas of Compute, and I think that architectural model is enabling us to actually look at different options in the cloud. Fantastic. So back to cloud health, that is cloud health enable organizations to move through their maturity journey. Yeah, I like to tell all perspective customers two things, which is, having done this for over a decade now, that you need two things to be successful in scaling in the cloud. The first is you need a superior product that's purpose built to drive and power your cloud center of excellence. And I'm sure you'll all forgive me if I'm somewhe biased and saying that I think that solution is is cloud health and is, you know, best and breed for that. But many people overlook the second thing that you need, which is you need an organization behind that product that can support you as you start from where...

...you are today in your maturity model and start to slide up into the right in your maturity and knows what you need today, nose, based on working with hundred or thousands of different customers, what you're going to need tomorrow, and is there to provide the people led services to enable you to be successful. and is why, you know, from the Second Watch perspective, I think it's particularly important, which is one of the reasons why cloud health, you know, really likes to partner with companies like Second Watch, is because, if you look at the complexity that's going on in the cloud, there is no way all customers can navigate that without finding a partner who really has, you know, focused in become an expert in the space. It's the only way you're going to be successful scaling in the cloud. And so so we strongly believe in that. We know a second watch strongly believes in it, and I think we come from a place of realizing that you can't go it alone and be successful in the plot today. That's great. Joe. I really want to thank you for your time this week. It's been a pleasure to get to know you a little bit and hear a little bit more about your story. It's very impressive. Rob Thanks for Co hosting this week and audience. We're always looking for your feedback. Please email as at cloud crunch at second watchcom with ideas and suggestions. Thank you very much. You've been listening to cloud crunch with Ian Willoughby and skip Berry. For more information, check out the block second watchcom company block or reach out to second watch on twitter.

In-Stream Audio Search

NEW

Search across all episodes within this podcast

Episodes (33)