6 - Hybrid Cloud

Posted on Sunday, May 10, 2020
Joining me in this episode is Thomas Maurer, a Senior Cloud Advocate at Microsoft who engages with the community and customers around the world! Most of the focus so far in Cloud with Chris has been on Public Cloud, so Thomas and I adjust course slightly, and talk about Hybrid Cloud, and the new and upcoming features of the Microsoft Azure platform. Get ready to be introduced to a wide variety of Azure Technologies, and how they may be able to strengthen your on-premises, public cloud or multi-cloud deployments. Let's listen in…

Show Notes

Hello, and welcome back to this episode of Cloud with Chris! You're with me Chris Reddington and we'll be talking about all things cloud.

The past few weeks and months have been a strange time for us all, so please continue to stay safe and healthy! I hope that these podcast episodes are providing you a little bit of variety, and something different in the day-to-day. Now, don't forget, that you can listen to these on your favourite platforms, whether that is CloudWithChris.com, Google Play Music, iTunes, PocketCasts, Spotify, Stitcher or YouTube!

So, let's start talking about episode #6! Joining me in this episode is Thomas Maurer, a Senior Cloud Advocate at Microsoft who engages with the community and customers around the world! Most of the focus so far in Cloud with Chris has been on Public Cloud, so Thomas and I adjust course slightly, and talk about Hybrid Cloud, and the new and upcoming features of the Microsoft Azure platform. Get ready to be introduced to a wide variety of Azure Technologies, and how they may be able to strengthen your on-premises, public cloud or multi-cloud deployments. Let's listen in…

Chris: Hi everyone and welcome back to another episode of cloud with Chris. Now we're continuing that discussion of cloud, but we're going to be thinking about a different type of cloud this time. Lots of us think about public cloud. Some of us may have heard about private cloud, but there's this also grey area in the middle called hybrid cloud, where we think about maybe our on premises. Or it may be even multi cloud for example could be another scenario of hybrid and it's one of those areas that I'd like to explore today. I'm very glad to be joined by another colleague at Microsoft at Thomas Maurer. So Thomas is one of their senior cloud advocates at Microsoft. He engages communities across the world and he's recently joined Azure Engineering beforehand. He actually was a lead architect and a Microsoft MVP and is very familiar with these kind of hybrid cloud architectures and these hybrid cloud deployments. So I'm delighted today to be joined by Thomas. Thomas - how you doing, Sir?

Thomas: I'm doing great. Hello Chris, and thank you very much for having me at your podcast. Really appreciate it.

Chris: No, very glad to have you here. Thank you for coming. So one of the things I want to dive straight into here. Starting to define what we mean by hybrid cloud because some people might be thinking of hybrid cloud as private and public. But there are - I guess other - scenarios, right - like multi cloud for example. So maybe if we start thinking about what you see come up?

Thomas: Oh yeah, absolutely. I mean, I, I really think that's great thing to start that. And yeah, if we talk a lot of people know about private cloud and their on Prem Environment. Then having the public cloud. There was this thing in between which we called hybrid cloud or which we still call hybrid cloud and this is absolutely still the case, right? This is absolutely still there. However, since - like - during the Ignite 2019 time frame. We changed our messaging a little bit in terms of that, hybrid is not just like something in between using on Prem and Azure. Hybrid can also be if you use multi cloud environments, right? So wherever you basically run your systems and your services this can be in your data center. It can be at the service provider. It can be like in your factories or your retail stores or at the edge or, again, another cloud provider, right? and I think that is really important for people to understand that. We are really going in that direction and we really hitting that as one state of doing cloud.

Chris: Right, so hybrid cloud is this idea that you might be running on Azure. But you also might be running wherever it makes sense for you to go and run your workload. We're not going to define where that is. You're going to define that based upon whatever your requirements are. So, if you've got - like you mentioned - some kind of manufacturing scenario, you might have some factories, some warehouses that have IoT devices and you need to run certain things on the edge very close to what you're doing. But, then you have some kind of processing that happens back in the cloud or maybe on premises or maybe even another cloud because you've got some kind of requirement that drives you across multiple clouds . And it's all part of that hybrid and multi cloud-idea there, understood. So, maybe let's pick up on that point of requirements. Are there certain themes, certain trends, certain requirements, maybe that you see across different organizations that you work with that drive them towards some of these technologies, maybe?

Thomas: Yeah, absolutely. So, I think one thing which is very important and a lot of customers hear us, or especially Microsoft and also other cloud vendors obviously talk a lot about cloud. But for me it was very important when Jason Zander (which is basically the engineering lead of Microsoft Azure) went on stage and one of the Microsoft ignite keynotes and said that we think that - or we believe that - hybrid is an end state for our customers, right? It's not in between state until everything is moved to the cloud. It really is like we see the need of customers having using the cloud taking advantage of the cloud and running a lot of services there. But, we really see that again that you have needs, right. For example, in a way that you run services in a factory, right? You don't want probably to have dependencies on an Internet connection to your cloud provider. And if the Internet goes down your whole factory stands still. So you want to probably be like, OK, hey, I'm running the factory, I'm running IoT workloads, running server workloads in that factory. And I'm sending my data on later to Azure; to the cloud to do like further processing or backups and whatever you do there. But I don't have a direct dependency on the cloud in that case. So that is definitely one of the large hybrid use cases we are seeing. The other one I would say I see a lot is that you actually are running on Prem, like you're an enterprise which runs workloads on Prem. You already have everything in place, but now you start to take advantage of certain cloud services right? So this can be, for example disaster recovery and backup where you can go out and just use Azure as a disaster recovery site for your data centre. Or that you store your long-term backups in Azure. And then we have a couple of other very interesting Management Services which you can use to basically make your On Prem environment - or again - I should say your hybrid environment - not just including on Prem, but also multi cloud environment, even better using Azure.

Chris: Nice, nice. And I've gotta put a call back there to one of the things you said in the beginning which was around the large hybrid use case. The direct dependencies, and not depending on Cloud for your hybrid workload. I'm going to say hybrid, not on premises, but let's go back to that factory scenario, and that idea that maybe you have some workload that if the connection between the factory and the data centre in Azure (or whatever cloud provider) goes down for example. Then you have that business continuity, you have that ongoing operations. That is so important, and I just really want to emphasize again that requirements really drive that decision, right? It's not just a case of “Oh, you're in this workload and you're in this industry and you're doing this thing” that suddenly XYZ is suddenly the right thing for you. You need to really understand your business needs. Understand what the damage potentially is for example, if that link goes down and just understand how the business can continue. That's so important, and I think we've spoken about one of those things in the previous podcast, which is this idea of a failure mode analysis. This idea of being able to look through the whole solution, not just the cloud aspect but end to end. You know whether it's in the factory, in the public cloud, in another cloud, or in the edge. If something fails, what is the impact of that?

Thomas: Oh yeah, absolutely. You should always look at your workloads, right? And you brought it up in a way that you have to look at it for, like Dependencies and like making sure that things are running. But, also in a way that you need to figure out… How can I make things better using the cloud, right? Or are things may be better if I don't run them in the cloud and maybe closer to my customers, right? There are definitely like different concepts. When I look at it, I basically think that if you very high level look at hybrid today - and you correct me if I'm missing there something. But, there are basically 2 two large concepts. One is basically connecting workloads which you run currently again on Prem or at the edge to the cloud and make them better. Or, basically bringing cloud services down to your environment on places where you can't run the cloud right? If you think about cruise ships with no Internet connectivity, or you have some regulations on compliance scenarios where you can't really run services outside of your country and you don't have an Azure Data Center in your country. Then they are also good reasons to have that. I think I think we are one of the only cloud providers, or if not, the only one who have as a serious solution there with Azure Stack. With our Azure Stack portfolio. But not limited to that, but like where we can actually allow you to run Azure services in your datacentre or at the location wherever you need them.

Chris: There's two pieces to that, and I think let's maybe start with the first aspect that you talked about, which is the idea of bringing cloud innovation to your on premises. You don't necessarily need to use cloud deployments and actually deploy things in the cloud to benefit from the cloud. When I think of the Azure Suite of technologies that is out there. Things like Azure backup absolutely work with an on premises or nearby environment, let's say to our users. But then also things like Azure update management, where maybe we need to go and manage some updates on machines for example. Or even some of the newer, shinier things that are coming out like Azure arc. Maybe let's speak about Azure Arc a bit later, or in a moment. But Azure Arc is one of these great technologies which is really, actually being driven because of the cloud and because of the nature of the scale that we typically need to operate in the cloud. But it now opens up this brand new opportunity for these environments that might be either on premises or indeed, like you say in other clouds as well.

Thomas: Yeah, so you're absolutely right there. The Azure arc piece and in connecting these services to Azure that gives a real - immediately - like a benefit to when you start using these right? It's not about just “Hey, we want you to move everything to the cloud”. No, no, no. We really want to make sure that you have the best solution. Whatever you need to do, and we understand that you need to run stuff On Prem and you can make it better using our services. I'm super excited about Azure Arc. In Azure we have this great thing called the Azure Resource manager, right? Azure Resource manager offers you all these cool things. These logical things like Azure Resource groups, tagging, stuff like RBAC and basically all of that right. And today we use that to organize and manage our deployments in Azure. So that was like all the Azure services. Basically if you interact with Azure you either use the portal, CLI, PowerShell or some APIs. But at the end you're interacting with that Azure Resource Manager. That one really take care of deploying these services and managing these services. If you have ever worked with that, you see the benefit of Azure Resource manager. And now the question really came up with like “Can we not just use these the Azure Resource manager to basically manage services and servers and Kubernetes clusters and a lot of other things which are do not run in Azure”, right? Why not extend the Azure Resource Manager to allow us to manage all the resources as well. And that is really what Azure Arc is. It basically extends the possibility of Azure Resource manager to manage services which run outside of Azure.

Chris: Understood. I'm equally excited about Azure arc as well. I think this idea of management at scale is really well understood, well attacked, I suppose with Azure Arc and it's really well taken as a problem. I love some of the demos that they've done. I think it was at the previous ignite actually, where they did this. They showed the whole idea of a Kubernetes cluster that you can stand up and maybe that Kubernetes cluster was at a retail bank for example. You might have hundreds of retail banks across the country, and to bring a new retail bank online all you need to do is just enrol the Kubernetes cluster with Azure Arc. It identifies that has the appropriate policies and what not and then basically, does itself like a self-install. And then you've got a new branch online and ready to go. And when you think of that power that that then enables, that's just awesome. It's really exciting when you think that technology can start unfolding some of these new scenarios.

Thomas: Yeah, yeah I really agree with that. Imagine. I'm sure you deal a lot with our customers when you're working in the FastTrack program where you have these customers which are doing this hybrid architecture, right? Where they say “Hey, I want like this application runs partly in the cloud, but partly also on premises” and I want to deploy at multiple times and very quickly. Now since you have everything connected to Azure in this Azure Resource Manager, you just go out, deploy an ARM Template, deploy the resources and you can set up that environment that application in one go, right? And it's not just about “OK, I need to deploy something in Azure, and I need to deploy something on Prem”. Now you can basically use the same mechanisms, or the same control. Let's use the word control for that to manage these resources. And absolutely the example you brought up there with the bank and connecting a new Kubernetes cluster. This is kind of like, mind blowing right? You can think about a retail store, where they run containerized apps. And then on the Kubernetes cluster, they basically have to take a hardware piece (and we can talk a little bit about what that hardware piece can be in a bit). But you take that hardware piece, you have a Kubernetes cluster on there and as soon as it comes online, it connects to Azure using Azure Arc. And it gets all the configuration policies down. It gets all the application policies down to that Kubernetes cluster and spins it up, right?

Chris: Absolutely. Just to be clear for anyone listening who has an understanding in Kubernetes, and is thinking “Hold on, I can do that with normal Kubernetes” because of manifests, helm charts, all these things are just transferable. The real benefits and the real value in what Azure Arc is doing, is having - as we've really been hinting here - this central management pane. This central management platform, and this ability to be able to just point a cluster to a specific configuration. What it's doing in the background is using GitOps and maybe that's something we can explore in another episode, for example. But it's using the idea of GitOps and allowing the cluster to register itself to a particular configuration store, which is like a repository of some sort. And then, because you're not having to push those changes out to brand new clusters or sign up new clusters and push them out through your DevOps pipeline. It just does it using GitOps. So, it's a very very clever way to help us scale some of that management which, Let's face it - sometimes the management side of things isn't the most fun of things to do from a technology perspective. You know, we just want to get things running so we can keep building the fun and shiny things. Keep the lights on and work on the biggest and greatest innovation that we're going for next.

Thomas: Yeah, absolutely. You absolutely nailed it with that. I think also for people who are never heard about Azure Arc. I think we now talked a lot about Kubernetes, but Azure Arc is not just for containers and Kubernetes, right? As of today, we can also connect just standard servers, right? These can be Linux servers, these can be windows servers. So you can also push down configuration policies basically to your servers wherever they are running. Using Azure guest configuration policies for example. One of the great things I saw, like one of the demos we're having and doing is figuring out “Hey, which servers are having insecure password settings”, right? We could use that policy before, for Azure Machines that are running - or servers running in Azure. But now, we can basically use them for servers running anywhere, right? And then the third part which was announced during Ignite as well, was the data services part. And that part is a little bit different from the other two. Basically, because with the Kubernetes part and the server part, you're basically connecting your existing resources or newly deployed resources to Azure. With the data services part, we basically give you the capabilities to take Azure SQL and Azure Postgres, to let you run that wherever you need it, right? So you get Azure services which you can put into your datacentre. Because a lot of customers told us “Hey, Azure SQL is amazing. It's great. We want to use it for our data. However, we have these scenarios” and I think we are coming back to what we discussed earlier… We run a factory or we have a place where there are regulations in place, where we can't really give it into an Azure Data Center because you're probably not in the same country or whatever the reason is. But this allows you just to take these Azure data services to run Azure SQL wherever you need it, right - and I think that that is also pretty cool.

Chris: Right! I guess I'm showing my hand a little bit here, talking so much on the Kubernetes side. I'm an App Dev at heart, so you know - I'll keep focusing on that! But you are completely right there, completely right. Virtual machines, the SQL side or the database side, absolutely - Those are other areas where you can focus on from Azure Arc as well. There was one thing in there that you mentioned which was about being able to put Things inside of your datacentre. You mentioned that from the database side with Azure Database for SQL or whichever service for example. Maybe let's pick up on that a bit more, because I guess Arc is just one way that you could potentially do that. There are plenty of other ways as well - because - and it's probably one that is more well known, I would say because it's been around a little bit longer. But, we've got this family of products, this suite almost - that revolves around this Azure Stack portfolio. So maybe, let's talk a bit more about Azure stack.

Thomas: Yeah, absolutely. Before we before we start talking about a stack, I just really quickly want to point out you're absolutely right. The thing with the Azure data services part. The way we get those Azure data services in Azure SQL Down is basically in containers. So our customers - if they want to run Azure data services using Azure arc, they will need to have a Kubernetes cluster somewhere and that's how we bring it down so that super exciting for you I guess. For me, as well.

Chris: Exactly! See, I told you. Kubernetes is the way. Kubernetes is the way - absolutely. Azure Stack - Slightly different to Azure Arc because Azure Arc - we've really been talking about bringing some of those management benefits that we've realised from public cloud. But Azure Stack really takes all of that one layer further, doesn't it?

Thomas: Yes, absolutely. So Azure stack really got the big big push as well. I mean, it's there for a while. I think the first time it was announced was back at ignite in 2015. I remember that - I think it was in Chicago where they first publicly talked about it and since then it really evolved. As you said, it kind of became this family of products. When we go back to that example where you basically have a new branch office or a factory, or the store we were talking about. You still need some hardware, right somewhere. That is where Azure stack the Azure Stack family comes in. Currently we have three different products or services If you will. We have the Azure stack hub which was renamed from the original Azure Stack and this one was renamed to the Azure Stack Hub. This scenario is really about bringing Azure into your on Prem environment, right? It really is a small - if you will - a small instance of Azure allowing you to run. Having an Azure portal. Having the Azure Resource manager. Having Azure services in your datacentre, and it enables a lot of different scenarios which we can speak in a bit. And then we have two other things. We have Azure stack HCI which is our hyper converged solution, which is basically validated hardware which we took to work together with our OEMs and you can go from very small two node clusters up to very powerful 16 node clusters. Then, obviously spread that over multiple clusters if you need to and run virtual machines right as you would do for that. And then the last one which was added to the Azure stack family was Azure Stack Edge. If you probably were familiar with, or the listeners here were familiar with Databox Edge. Basically, we renamed data box edge to Azure stack edge and announced a couple of new capabilities there back at Ignite. And what it does, it's a first party server unit which you can order from Azure using the Pay as you Go model and you put it in wherever you need it. And, it runs containers - there it is again, right? So you can run Kubernetes on top of it. Well, that was one of the announcements which we made there, and this enables you to run AI and Machine Learning services right at the edge, right? It comes with super-fast storage, so think about when you have collected large amount of data and you push that to Azure Stack Edge. And then, you run container like services inside containers to analyse these using machine learning and then later on you basically just replicate it to Azure for example as the long-term storage. So we have like really cool products there as well.

Chris: You're making my job as an interviewer both really easy and really hard, because there's so many things that you just said there! I'm thinking I want to explore that. Let's talk about that. Let's talk about this. Let's first talk about one of the things you mentioned there, with Azure Stack - as a portfolio - really what you're doing is you are deploying a small instance of Azure is how you worded it. And I completely agree. Because when you look at it, even things like Azure Resource manager templates - they still translate and you can deploy using the same approach that you use, no matter whether it's Azure; Azure Commercial, whether it's Azure Gov or one of those sovereign clouds, or even your own deployment of Azure Stack as well. You've got that same translatable, same transferable deployment approach as well.

Thomas: Yeah, exactly, and this is exactly where we put it in, right - where I've seen customers doing this for different purposes. One which I called the technical use cases, when the Internet connectivity is just not there or it's just not good enough and you have too much latency for the application to run in an Azure Data Center. You can use an Azure stack hub and bring it closer to you, your users or your application or wherever you need it right. The second one really is about regulations and being aware of that, so some companies, they have regulations that they can store data outside of their country or even outside of the datacentre. And so, this allows them basically like a solution for that and still run not just VMS. Right? We're not talking about the big benefit for Azure Stack Hub is not that it can run just VMs. It can run app services, it can run other Azure Services as well. Well you have a similar deployment model using the Azure Resource manager as you just mentioned. And then the third one really where I've seen - and I never really believed in that one until I saw it and I heard it from so many customers - was that cloud modernization, right? Customers who said, “OK, hey, we want to go to the cloud and we want to take advantage of the cloud deployment model and we want to take advantage of these cloud services and PaaS and all of that and doing infrastructure as code and all. But we are not ready yet”. We don't have like so many things not in place to use the public cloud. But we want to do that. And then, Azure stack. Basically Azure Stack Hub gives them the same thing as they would run in Azure, but without actually leaving their datacentre, right? Really running Azure under their terms and I think those three reasons are already good reasons for Azure stack Hub.

Chris: Absolutely. It's that dipping the toe in the water isn't it? It's preparing for that potential cloud move later on. Or even in fact, you know what? You don't need to move later on - you can carry on running where you are because you're still getting the same benefits. You're still using the same deployment models. And if you do decide to go to public cloud a little bit later, you know - requirements allow you later on - then, great! Just use the same deployment model, use the same concepts that you've learned and deploy into the public cloud. Lovely idea, lovely concept.

Thomas: I also had - I mean that goes into like one hand… I was talking to one of our customers, recently. I think it was before I joined Microsoft. The requirement they had, they said - “Hey look, we're going to move all of our websites into Azure and we are absolutely fine with that. However, we have this internal regulation. So what we have is a product which we launch one time a year and we have this internal regulation which says - hey nothing about that product can leave our facility. Not even like… It's just a rule.” You can tell them like Hey Azure is super secure and everything. They don't. They really don't care about that. It's really just about “There is one rule and nothing about the product leaves the facility”. So they said, “OK, hey, we're going to prepare that launch website in our data Center, but we want to have the same deployment methods or you want to use the same thing as. We then would run in Azure, right? Because in Azure, we're using PaaS services. We are taking advantage of app service and stuff like that. So we also need that run in our datacentre and so they could.” They take basically Azure stack up and they're working on their secret project and their launch website on Azure stack hub. And as soon as they are ready to publish that, they basically publish it in Azure at the specific date, so that's another. When I heard about that, I was like OK that that really makes a lot of sense

Chris: Gotcha. Nice, yeah - that's another interesting use case. Excellent, excellent. So let's maybe expand into some of the other areas. We've talked about Hub a little bit there. Azure Stack HCI - We mentioned about the ability to run containers and things on there as well, and have some edge compute there. And, I love that idea again… Why do we need to worry about sending things up to the cloud when we can maybe do some pre-computation of things locally, where it's closer and get that response back so much quicker. Love that idea.

Thomas: Yeah, absolutely. I mean this is - this is really like the key, right? And it's not just about like a nice to have option, it's also in some scenarios it's really critical right? We were discussing about like if you do facial recognition and things like that. Then you sometimes really need to have a fast response, right? Or if you like measuring something in the factory floor, you don't want to spend, like 10, 20 seconds basically getting an answer back from the cloud. You want to react really, really quickly, and that is exactly what we are allowed to do with Azure Stack Edge.

Chris: Absolutely. It's almost thinking about those potential health and safety scenarios, right? Whether there's a security issue, or whether there is like you mentioned in the warehouse like a spillage or some kind of accident. Those extra seconds could be crucial - could mean life or death. So having that ability to deploy again, that same model in the same way, closer to where your users are and where the actual application is. That's an amazing offering. An amazing idea. So, love that.

Thomas: Yeah absolutely. And as you said, now the great thing here - if you think about it… We have these hardware appliances, right? We have Azure Stack Hub. We have Azure Stack HCI. We have Azure Stack Edge and now. And now, if you think back - we talked about Azure Arc. If you combine those things together, you get basically an amazing solution. So think about - you put in your Azure Stack Edge in your store. In your factory, you spin up the Kubernetes clusters. It connects up using Azure arc, it connects to Azure, connects to the policies and basically gets down these policies and configures itself and deploys the application. As you said, using like GitOps and things like that. It's just like it gives us so much power, right?

Chris: Absolutely. I think the keyword that keeps coming back here is scale, isn't it? It's the management of scale idea with Azure Arc. That deployment at scale idea, whether it's on public cloud or whether it's on something like Azure Stack Hu or whatever product from the Azure stack portfolio. There's this idea of being able to have this universal deployment plane, and this universal management plane. It's really exciting because I think so many times in industry, we learn one thing and then we have to relearn it, and re learn it because that model has slightly changed overtime. New technology has come in. But what we're doing is saying “Actually, this was good enough for public cloud. This is good enough for us to do in this hybrid scenario as well. So let's bring the best of that, bring it back and reinvest it in this new scenario as well”. And I think, one of the areas, maybe we've neglected a little bit while talking here, but is also things like the IoT side of things. We've kept talking about edge, and there's also this concept of IoT edge as well. And once again, this similar theme that seems to be keep cropping back up, is this idea of containers. IoT edge once again works in the exact same way . This idea, that you can go ahead deploy containers to those particular IoT devices running IoT edge. And again, run it literally on the device. When you start thinking of all these different offerings that we have talked about, whether it's IoT edge, whether it's Azure, Azure Stack HCI, Azure Stack Hub, or even public cloud. There's so many points there, then that you can start deploying your workloads and it gets quite interesting and exciting - the different use cases you can start unlocking there.

Thomas: Yeah, absolutely. I can't stress this enough. And I think, what also is very important - what we have to realize is it's just there to enable new scenarios. You can say, “Hey, we could do things like this before without the cloud, right?". Yes, probably. But think about the complexity you were building up, just building all those network connectivity things you needed to build, right? For example, you had to add VPNs or whatever you had to like from all your branch offices or stores or whatever. I remember that, I had a customer once where they connected all their retail stores basically back. I don't know what they managed like 200 VPN connections… which again probably there are people out managing much much more. But, that was already big effort out there. And to build just that base infrastructure to make this all happening, they spent a ton of manpower and resources. And now with that moved to the cloud. It really brings us to a higher level, where we can as - in the case of - IT administrators or IT engineers or developers. We can basically focus on what really matters. We can start improving the application experience or application itself. I think, get an easier management and focus more on things like security as well, right? I think that is also very important because when I think a couple of years back, I always said “Hey, people are not really thinking about security. They are happy when things are running, right? They are just happy when the application or the platform works. but they didn't even spend that time to build a secure infrastructure or environment”. And now… Now basically we're giving them “Hey, here is the thing. It's easy to set up. It's easy to connect. It's basically the base. It's secure, designed by default. But now you can focus on security on your environment or on your application”. I think that that's a huge benefit.

Chris: Yeah, a ton that people can be listening in and thinking OK… These are different options for me to consider here. So, maybe to start wrapping us up here, let's think about some of those people who maybe are at the beginning of their journey. Whether that's beginning of cloud, or listening to this and thinking “Actually, I've got a really great use case for hybrid cloud”. We've talked about things like site recovery, backup and update management, for example, where you can go ahead and use those with your on premises environments or other clouds, or public cloud, wherever that may be. We've talked about things like Azure stack portfolio, so things like Azure Stack Hub, HCI and edge where you can actually go and deploy a cloud in your own environment. And things like Azure arc where you can scale that management side of things as well. So bringing that all together, what would you say? Or maybe some of the top things that should be on these people's minds as well, and maybe some of the tips and tricks and some of the observations that you've had where you've worked with other organizations along that journey?

Thomas: Yeah. As from our podcast right now, you can imagine there is so much going on and so hard to basically say everyone here's what to do. But if I can recommend, what I would do if I start seeing the opportunities and the things we have here, and we've talked about here. I would recommend start small, right? Don't go and try to say “Hey, look I want to change everything we are doing today”. Because if you do that, you build this large scale project and you have so many dependencies and you're more likely to fail. My tip really is start small. Start with whatever is top in your list, like for example managing updates for your servers across your environment. In that case, have a look at Azure Update Management. Start having a look at it and start implementing that and see how it works and what the benefit will be. This can also be just for backup or whatever you need, right? We talked about Site Recovery as well. Start with something you can actually handle, and you can actually get a benefit very very quickly. Then you can see - “Hey, that worked great and you will learn a lot about it”. You will learn how to work with the cloud and how to basically interact. You will see what you need to think of, and you will learn that. Then you can move on to the next thing, right? Don't go out and just - I mean you can create that vision in your head and say - “Hey, that would be the perfect world if we would do this”. But, let's start small by just doing one thing and learning from this. I think that would be my tip.

Chris: Excellent. And, I have to say as well… I'm sensing lots of themes here at a line from a DevOps perspective as well. A few episodes ago, we had Abel Wang come in and talk about his DevOps journey, I suppose. There was something that you mentioned earlier about allowing people to focus on growing their applications. Building better applications. Being able to focus on security and these other things. And I think it's that idea of delivering value to end users. That idea of starting small. Working in an agile away, iterating over that. Building, growing learning, and being able to deliver that value to end users. There's so many similarities and synergies there from a DevOps perspective as well. So, really great to hear that guidance. Really great to hear that feedback. Thomas, I think there's absolutely plenty of information in there for folks listening to go and dwell over or think about. So I'm sure, no doubt that we'll have to discuss this in a bit more detail… Maybe some of the things that evolve over time with some of these services, and maybe jump into a little more depth. Perhaps in some of those industries, and some of those scenarios that we've talked about this. So Thomas a big thank you for me. Thank you for joining the show today. It's been a great pleasure having you here.

Thomas: Thank you very much, Chris. And we should definitely do that.

Wow! I mentioned in the intro that you would be introduced to a variety of Azure Technologies that may be relevant to on-premises, cloud or multi-cloud deployments - and I think that was well and truly covered! A big thank you again to Thomas for joining and sharing his brilliant knowledge in this area!

I would love to hear how you're finding these episodes, and if I can make these any more valuable for you. Please do get in touch either on Facebook or Twitter, @CloudWithChris if you have any feedback! And as always, if you enjoyed this episode - please don't forget to provide a like or subscribe on whichever platform you have been listening in on. If you think it might be useful for your own followers or connections, please do share onwards!

Thanks once again for listening in. And, until next time - Goodbye!

Guests

Thomas Maurer

Thomas Maurer

Thomas works as a Senior Cloud Advocate at Microsoft. He engages with the community and customers around the world to share his knowledge and collect feedback to improve the Azure cloud platform. Prior to joining the Azure engineering team (Cloud + AI), Thomas was a Lead Architect and Microsoft MVP, to help architect, implement and promote Microsoft cloud technology.

If you want to know more about Thomas, check out his blog: www.thomasmaurer.ch and Twitter: www.twitter.com/thomasmaurer.

Hosts

Chris Reddington

Chris Reddington

Welsh Tech Geek, Cloud Advocate, Musical Theatre Enthusiast and Improving Improviser!

Chris is currently a Senior Engineer on Microsoft's FastTrack for Azure team.