Cloud Crunch
Cloud Crunch

Episode · 2 years ago

S1E04: 5 of the largest lessons learned over a decade

ABOUT THIS EPISODE

We analyze the 5 most important lessons we've learned over the past decade from the point of view of how they affect a new cloud user as well as an advanced user or organization. We dive into governance, optimization, CCOE, mutli-cloud, talent, the knowledge gap, breaking down silos, scaling services, and the shared responsiblity model.

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, Jeff Aiden, CO founder and Executive Vice President of marketing and Business Development, Ian will be chief architect cloud solutions, and Skip Berry, executive director of cloud enablement and now here, are your hosts of cloud crunch. Hey, everybody. Jeff Aden, co FRINDER, Second Watch here, joined by in Willerby and skip Berry. Again, thanks for listening in. Today we're going to recap the five most important lessons we've learned over a decade for new users. That we're also going to add in five lessons learn of advanced users and organizations or early adopters. Hey, guys, let's start with the advanced ones today and talk about some of those lessons learned. Skip yeah, so going through the list here, if you will, I think we should talk about governance as a big area, optimization, another cloud center of Excellence, CCOE, multi cloud, and then just general talent. Right. How do you keep your talent up the snuff and continuously the sharpen the knife's edge so will CAREC US off with governance. You know, customers that are outside of, you know, POC single application. They have hundreds, if not thousands, of accounts. What are some of their governance challenges that we're helping solve on a daily basis? Yeah, I think the big one really is making sure that they're not doing things. You know, you talk about shadow it is one area doing things in a silo. Right. This is about moving the business into a direction where everyone has visibility into, you know, the best practice as the involvement of the business, leveraging lower operational costs and then adding top line growth. So while a lot of times, from a governance perspective, you know, we follow change management, etc. This is about evolving the business into understanding right, the critical role that I plays now as adding value to the business. So that's part of the governance that I think needs to be embraced in the advanced users space. And what are some of the tools that you're helping or automation that your team's helping customers leverage in order to make sure that their governance is, you know, up to the latest innovation or, let alone, you know, up to some of the more recent risks. I think there's certainly some areas that haven't changed. We'll talk about those. First is you always need a good tagging strategy, no matter what Claud you're in. That is first and foremost. It is a fundamental pay to play strategy that you must have. Once you have that in place, one of the biggest thing, like governance, is kind of normal evolution, is you launch it with shadowy tea, a skip was talking about, and then people have administrative access to pretty much do anything they want to do. If they don't tag things correctly, they spin up resources nobody knows who these owns them, and they just stay in and run and they continue to accumulate expenses for the organization. Putting in the proper guard rails and place for this, such as a tows can fig, is a big way of handling that. That will make sure that when you build those business logic rules around it, and there's a lot of them out of the box that are just just work, that you will not spend up resources that somebody cannot figure out who won't it could shut them down automatically if the configurations are not done correctly, and that's really a big deal to make sure that everybody's following the rules, because there's a competition between governance and going fast. So you want to use as much automation as possible, and there's lots of it out there now, in order to accomplish that. Well, tell me, you brought a one that we use every day, right, the tagging strategy. But for larger organizations that can be difficult, right, for them to implement that across the board.

It is, and if you don't start from day one, it's a lot harder to come back and retroactively fix that. It could take months and up to a year in some cases that we've seen. So starting with that good foundation and making sure that it stays in place with the tools that will ensure that you not have to it's it doesn't become a heavy lifting item going forward. Well, it leads into our next topic, right, skip around optimization, because without knowing the tagging strategy, you can't build back to the project. You don't know who's leaving or from resources, who's using what. Right. So kind of leads into optimization. Is as a as something we see advanced users struggle with because they often just think it's our eyes. Right. Yeah, that is it's a well defined policy still must exist as far as like, you know, life cycle management from whether it's application, deployment or what have you, and then really about just having a good control where you spend, knowing what you're you know, basically everyone wants the greatest and latest beast out there to go power on, my biggest em that I have right, but having policy set forward and knowing what those builds are. Back to whether it's aws and having your ect builder all shopper and Shiny and know what is out there for inventory, or just making sure that we have, you know, established guidelines as far as you know when you doing a migration for Claud Readiness Assessment, that you know where the application is going to fit into. So those are big from an optimization perspective. And then really have the discipline to go along and we're necessary, give haircuts and make sure that budgets stay on point. Yeah, we don't want to give away all our secret sauce here on optimization, right, because we're really experts. But we do see this come up with large companies, especially as they start to engage with these enterprise type discounts, right, and you know what we've seen over the years. Is the simple band aid is our eyes, right, but at the end of the day there's more ways to leverage cloud native solutions and cloud native products to optimize, as well as just, you know, new classes of service, right guys? That then leveraging resources when they're appropriate. Right. It's not just the financial piece, right. Yes, not the financial piece all the time. You got to look at test V to usually wear a lot of the savings come to play, and it's not a onetime exercise either. It's an ongoing exercise that needs to be done on a regular basis. It's kind of like cleaning out the garage. Once you do it the major lifting, there still got to keep it up the date so as no services come out and as the skills increase with organizations, they have the opportunity to also not only just optimize their sped but optimize the applications themselves by leveraging more cloud native services. That will continue to drive the operating cost down. I think it's about learning new habits. Yeah, to you know, learning new habits in the cloud that you used to do on Prem is really this is a new habit. Right, to be continuously looking at areas that you can optimize really at the end of the day, and even though you clean out their garage, there's always some new toys to put in there right. Well, who's responsible for this? You know, often it's the cloud center of excellence, which is within the Organ Uization. At this point. Of more mature organizations, they would take on that or leverage a partner to work with, like Second Watch and other software tools. But let's talk about the clouds center of excellence and what we see as some of the best practices of the more advanced users on what that looks like from you know, not just the the people associated with it, but the rules and responsibilities. So we do encourage all organizations to form a class center of excellence or a CCOE. There's also another name that you may hear out there. It's not as used frequently, but it's cloud enablement engine. It accomplishes the same tasks. It needs to be a multi stakeholder organization, so everything from the business to finance, two operations to development. You know, it goes on and on and on. Everybody needs to have a representation in that from the art organization.

It's not just formed at the beginning and just once you get the cloud. It's done, but it's somebody, it's a body that will continue to revisit the new things that come out in the cloud and update the standards to make sure that those are enforced. The biggest part two is also just evangelizing within the organization to make sure that people are embracing the cloud, celebrating the successes, and that can be from the developers to implementing develops with cloud native tools, to optimization, cloud refactoring and all those types of things. And it they have to have multi skills as well and not everybody has to have every skill. Of course, on that on that organization, but as far as multicloud, hybrid networking, security, of course, all those types of things, and it's really sharing those lessons learned too, because everybody's new to this environment and it could be paralyzing. So if they do it together, they're not siload, things tend to go more rapidly yet and they typically have an executive that leads the center of Excellence and, as you point it out right, it's a collaborative cross function team that is subject matter experts in particular areas within the organization, but then also becomes subject matter experts in cloud services and cloud solutions. Yeah, and that's where, like, from my perspective, on the delivery side, is really about, you know, when we have the established team or how the customer, client is going to move forward. It's really instilling those values of, you know, being bold, challenged the norm, look about new ways how to do this. is about evolving the company to operate a new model. Right. So one of the big things that is always out there is, once you have the you know, the team established and we're in the throws of the delivery is make sure that we celebrate some really some early wins of success with the with the foundation of the CCOE that's in place, and then really about making sure the customer has the confidence to evolve from there and really that's their new norm of doing business. Yeah, and with more advanced users and organizations with the COOE's right. They've grown up with a lot of understanding and background with Amazon and with the evolution of the cloud. Here comes multi cloud, our next topic. Right, they have learned all the acronyms for a tows, they understand the products and services and now they're being asked to integrate another cloud, which may or may not be similar from everything we've talked about from a governance and optimization standpoint in order to deliver services within the organization. So what kind of challenges are you seeing today with those customers taking that base of knowledge and transitioning to Azure Google? Well, it is there's a little bit of a learning curve. Definitely. One of the things that we do tell people to do is is training, of course, is extremely important to this. The other cloud providers have training based upon their knowledge of AWS. So if you look at GCP, there's a GCP training for any of US professionals and Microsoft asure has the same thing. That's a great place to get started. It's really normalclature at the end of the day, and of course there's a little bit of technology differences as well, but ultimately a lot of the legos are the same legos. Maybe their different color, but they accomplish the same things. Understand that multicloud environment. It has become a challenge because it could become paralyzing of determining where to put things when you're a multi cloud environment. Typically what we're seeing as customers will go deep with one cloud and that tends to be the one where most things go to live and that is for financial reasons in a lot of cases, and it's also because of skills. The larger discount. You can get more of a discount when you push more things into one particular cloud. YEA, and that a side to debtail off of that really is when...

...we look at it, what is the best outcome here? Right? So sometimes solutions that the client is looking to do fits better in one of the clouds better than the other and then taken into consideration, as he and said, skill sets and what have you. You know, that's really where we drive and help educate the customer, you know, in light of two from a financial perspective of spend and whatever program they have. Those the things to consider. But one of the other big things that are almost as a tax, probably the wrong word, but almost invisible. But there is the culture. They may have grown up on an aws platform and what have you. You know, Azure or GCP is come into play now or whatever reason, and it's about, you know, breaking down those you know, our comfort areas as well to go and pursue that. This may be better for the business to go down a different path than what their customed to, and that's something from a second watch perspective that you know we're willing, able and happy to do with clients and helping them show that you know it's going to be all right and we're going to get through this and your your company is going to prosper from perhaps looking at a different avenue. Yep, and you know, listeners, you can eat google. Hirst Aws asure probably one of the more advanced users of multicloud, which makes sense. We've worked with them at a number of companies, a lot of their team where they previously worked, and you know they have many, many years of experience, which kind of leads to talent, right, which is our last topic for the advanced users. You know, we see talent being tight in the labor market, so companies have to do some exciting new things to attract some of the top talent. Right. How else are customers solving for the talent issue? I think typically customers want to see the talent develop internally. They usually have some people that are very strong technically, maybe new to cloud. So there's a couple different paths that people can do. They can really rely on the cloud providers themselves. They have quite extensive training that they could do with the client. They can go to classrooms, but that's great and all until you have to go use it. So part of it is is that incentive programs as well. So there's a lot of certifications available for all these cloud providers and incentivizing employees to go out there and get those certifications. It does make a big difference. And then, in addition, as we talked about with a CECOE, is really celebrating those early wins and evangelizing those because success tends to breed more success. Yep, different here from my perspective, advantage point also, it is just, you know, I'll look at it almost make a more's law contrast. You know, from technology, this is really more from skills. Now everything that used to be, you know, a year, your skills held up of whatever. Now it seems to be like three months, six months, even even more so, you know, from a client perspective, very empathetic of where they are, but you know, overall in the industry it's going to be a process. We have to keep on retraining, relearning and you know, this is the new norm that we live in today and it's a continuous pursuit to make sure that we stay on the edge and you know, I'm happy to help clients get there from our perspective. So again, it's not just about implementation and what have you. It's about bringing customers along the journey and how they evolved and show them how to recruit, even in some particular state circumstances. We've been working with customers for job descriptions, which seems so nonnorm from a Cot enablement, but at the end of the day it is helping our client evolved right in the space. So anyway, right, and you have brought up great points. And what we've also seen is those individuals that are involved in the CCOE Cloud Center of excellence that leverage the ability to learn outside their expertise. So if they're networking or their storage or, you know, whatever their expertise has been previously an old it, they start to understand that they need to understand a breathless services right without...

...the within the cloud in order to advance their careers. And we were just counting yesterday that I think there's over twenty five individuals that we've worked with that have been promoted by leveraging the clouds. So we've actually seen the inverse for some of the new adopters that will talk about next that fear that their jobs may go away. We've actually seen the inverse work. Customers that have leveraged and and clients that we've worked with that of leverage the cloud have actually been promoted in within their company as well as within others. So great recap on the advanced users. Let's jump over to to new listeners, so people that maybe leveraging one or two work cloads in the cloud and just getting started for a variety of reasons. But let's talk about some of the five lessons learned over the years that we keep seeing kind of repeat themselves. All Right, I think that I'm going to add as six. Then we'll call it zero, and it's security, because everything has to start with security and if you do not do it now, you're going to pay for it later. Talk about building up technical debts. So well, add that as the bonus one. But the first one to go back to the knowledge gap. It's that you really have to solve that issue. And again it's with incentives. It's finding that the people that want to do this and as an individual in the IT market, you become extremely marketable if you develop these skill sets. For that alone, you should jump on board and get trained, get certified and really learn how to dig in on this. And the knowledge gap is wider than ten years. At the rate of innovation some of these services that have build on each other over the years has just increased almost the time frame to catch up. But in other ways the innovation has reduced the complexity right. So a lot of things that we're done manually or through automation in the past they're now tools. But at the end of the day they need to start at certification level and work their way up in order to get a broader understanding of the services. Yeah, that's also experimentation. You have to have that environment and culture of experimentation. You can fail fast, way faster than you've ever failed before, strap back up and keep going. So, you know, create that environment where people can get in and try things, replicate things that they've done on prime in the cloud and really get their hands dirty. It's like going from a bicycle to a Ferrari right. So skip. You know, you lead a big team of pro services and obviously some of that talents grown within. What are some of the ways you know here it's second watch that we continue to get our associates challenge and learning more. Yeah, it's a great question. It's a question. Then I'm out by I'll say, executive management. I had many of our clients asked a lot and it really it's like the ladder of successes. You know, if you haven't done anything yet, the associate level of any program from any of the providers is always a great place to start, and then you advance up to the intermediate and the advance, or the professional as they call it, in each in each space. And then really it's about, you know, wherever you're natural inclination is or your curiosity is, because as human beings, we know that's where you know you pick up passion from. So you know things like devops security and then you know whether it's data engineering, whether they analytics or what have you. Those are some more finite places, but at the end of the day it's about exposure, you know, and that's really what we do with our folks right so look at the aptitude, what the background is in a lot of our clients of people and infrastructure would have you have had that great exposure and now it's about encouraging them on. You know, what their path is success could look like as their company evolves. So one that kind of leads into the next one, right, is how to break down the silos of subject matter experts and heard of that's created a, you know, cloud center of Excellence, which we've just talked about. But as they start to get this learning, maybe they've been a networking engineer, right, and as they started advancing through this training on the cloud, what's a little bit...

...different is they start to leverage areas and specialties outside of their expertise, but leverage them in a much faster way than traditional services. Rightly, yes, for sure. Yeah, it's a it's building the trust model all over again to write in those places where as you kind of Desilo, right, it's amazing. It's amazing how people still want to hang on into their little the worlds that they were before, not to actually I shouldn't say a little, you know. So it's about making sure that they continuously understand that they're always not going to be the expert. They maybe another expert in another business unit, in another line of business that's you know, whether there are stem from CCOE or weather, you know, wherever it is in the business, they don't need to be the sime in the whole area. Right. It's about picking a part of the business that they are smart on and then, you know, working together in an iterative fashion that the cloud actually promotes and getting things done. So, yeah, any and the next one on our list is scaling services, which you know, could be pretty broad. It's more about assembling solutions right, because cloud is not always cheaper. If you do it incorrectly, it could cause you some speeding tickets because you're going so fast. Right. So let's talk about some of the challenges for new adopters that are looking at adopting cloud services. Yeah, particularly if you've come from a monolithic architectural environment, it's very normal to say I'm just going to take that and do exactly the same thing in the clad. We see this with big data solutions, we see this with web applications and what have you. But understanding how to maybe break apart that architecture just a little bit, make it stateless, understanding that things do not need to be persistent and continue to just allow it to auto grow shrink. You know, we can take things like big data jobs, etl jobs that run once a day. A lot of times enterprises have these large clusters of to do about there running twenty four hours a day. That's no longer need it. If you need it for an hour, spend it up for an hour. You got the data. It's INN sthree or it's you know and other club provider storage areas. Run it. You can do all this automated shut it down, if it's successful, push the data into the next cycle of whatever they're going to do in the data processing. Same thing with web applications. Taking that and saying I'm going to run it in the you know, nobody's really hitting my new site at three am, so I can scale that down. But if there's a huge new cycle and cut something comes up in a lot of scale and it just understanding how to do those simple tweaks to the architecture. Then if you take it the next step further, you look at containerization and serverlist models, where you get a lot of those things and they're almost automated. You. If you look at Atbs Lambda and you're getting a lot of calls behind it, it's going to fire up and adjust it's Alve it to whatever you need you have. Those are great examples and one I think it kind of glossed over was around big data, and I'm going to challenge you on that, right, because we've seen this time and time again. You know, the traditional way it services have been leverage, right, as there's a business case in the business case is some future use case, right, and a customer will either use assets within their data center or provision more resources for a future state. So you have all this unused capacity. We've seen examples where customers actually take that business case and try to apply it to the cloud, not understanding scalability. So, Hey, someday they might need two terabytes a data that's what their provision in the cloud. However, today they only need two gigs, right, and they end up over provisioning and saying it doesn't cost less. But if you look at that over five years and the business case actually got to terabytes, oh you're going to save money, but if it fails, you just turn it off and start over, right. I was going...

...to drive into that point exactly, Jeff, about especially new users. Right, this is almost in a timidating, you know, frozen mode for that they can't get out of their own way because they're still used to planning infrastructure and have you know, buying cycle and budget etc. And this is such a revolutionary idea for them to wrap their minds around. But that's a perfect example of, you know, an area where we can help a second watch, to help them plan that out for them and understand that scaling. Actually, yeah, it's not only mapping that out right, but it leads into our next point of putting services together. Right. So, the old days it was pretty much easy to and has three and then all these services glacier, I mean, you know, now new databases etc. Talk about, you know, machine learning and AI. Right, let's talk about putting services together as a new user. Yeah, I think, like all good stories and everything that we've heard, it's the outcome. What is the outcome that we want, right? Is it a is it an upgrade, you know again as a lift and shift? Is it just, you know, totally advancing the business forward to be more agile or more innovative or what have you? And then we look back and, depending on what cloud service provider we have, we will go through the you know, whole assessment phase and then plan it out on how we will get there. Many services, almost too many of the list if you want to go through that, but basically, you know, it's your foundational builds all the way through, I'll say, you're up the ladder, if you will, on enhancing that foundation of bill all the way to putting the application on and making sure whether it's a scaling application needing to scale or what have you, and then through your running optimization continuously on it, I would have you. So those are just some very high level, broad based putting services together for a solution. I don't know even if you want to speak to anything more specifically. Yeah, I mean absolutely the the other option to is to take some of your monolithic approach as well before and start breaking a part, and that's it, just simple things before you can add a load balancer in the front end of it, at a Waff in front of it, all kinds of just simple little building blocks that you put in front of it. Ultimately your price is going to go down, your performance is going to go up in your operational post will go down as well as far as the level of effort, but also use automation associated behind it, as far as the operational aspects again, autoscaling, those types of things, and just to allow the cloud to take care of itself as much as it possibly can with a lot of these newer services well, and that leads into our third point, the shared responsibility model. You kind of started out with Zeros security. You know, it's not just the cloud provider takes care of it. It's not just the company takes care of it or the partner. Right. There's a shared responsibility in moving to the cloud and it kind of goes through the governance from the advanced users. But let's just talk for a moment about the shared responsibility model, what it is and how customers can understand it a little bit better. Absolutely so. In the past, if you're building at your data centers, you were responsible for physical security, starting there to making sure you destructed drives correctly, data retention, all those types of things, and then just employee background checks. Now that you're not even stepping foot into a data center, models significantly different. So Amazon handles a lot and so does Azure and GCP to handle all these things that you used to really have to do. That we're burnsome they're not anymore, but that doesn't get you off the hook. You're still responsible for making sure that you're using a cryption and transit cryption at rest. They give you all the tools to do it and they certify it with auditors, but you're still your responsibility to architect your applications correctly and make sure that your data is put in the areas with the proper roles and responsibilities for access. One of the bigger things that I've seen that is a misnomer behind security in the cloud is will say, Hey, I understand that this cloud provider has PCI compliance, so I'm just going to put everything in the cloud and PCI. Well,...

...that's not true. Their components are PCI and they can give you the adda stations behind it, but you're still responsible for your application of the design, implementation in the way you operate it in order to be secured. Key point to eat in as you still have to prove compliancy on your end. Right. So absolutely, yeah, you're getting fifty percent of it already done, but the other fifty percent is up to you. Well, new users will actually sometimes ask to go see Amazons data center. Right, yeah, don't want to go see it and you're not going to. They're never gonna let you in there. Now you're going to tell you what they are, right, and I think that's that's good for the new users to understand is that, you know, these sites are calm black sites, but but they're not known to public. Your customer data is not known to the Amazon employee that's working within that a Z or that region. So they can't tip over a rack like your employee could write discerneled employee. They're not going to know that your customer data is on that machine versus that machine, or in this building versus that building. And they really have done a great job over last, you know, over a decade, in ensuring that that process is highly secure and safe for customers. Yeah, and absolutely, because they also bring in the auditors from all these different compliance areas to really make sure that they're following their own rules and they'll provide you these reports, all them well to so that you can feel assured that at least they're part of the shared responsibility model. Is Up to snuff and you're ready to go. Yep, Hey, guys, always a pleasure chat with it. I think this is really good. We got to keep it under thirty minutes for our listeners, both advanced and new. Thanks for listening in and send any comments to cloud crunch at suck watchcom and we'll talk to you next week. You've been listening to cloud crunch with Jeff aiden, 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)