PurePerformance - 047 101 Series: OpenShift with Martin Etmajer

Episode Date: October 23, 2017

If you believe OpenStack and OpenShift are pretty much the same thing. you better listen to this episode with Martin Etmajer ( https://twitter.com/metmajer ). He explains what OpenShift is, how it differentiates from Cloud Foundry and other PaaS platforms, and which major contribution it can have to successful DevOps transformations.To put it in his words: OpenShift provides great user experience for developers to push their code changes automatically, packaged as containers, into different environments without having to worry about where and how these containers run or how they scale up & down. You should also check out his presentations from Red Hat Summit on Monitoring and Logging in OpenShift ( https://www.slideshare.net/martinetmajer ) as well as more material on http://www.dynatrace.com/openshift.

It's time for Pure Performance. Get your stopwatches ready. It's time of Pure Performance. It's very special today because my co-host, the gracious one and only Andy Grabner is no longer up for grabs. That's right. He got married this weekend. Congratulations, Andy. Thank you. stronger up for grabs that's right he got married this weekend congratulations andy thank you yeah so for all of our fans who are like oh boy i'd really like to meet andy because i heard he's a great dancer and all that you know what sorry he's now officially taken well i'm still dancing right i'm still dancing with other people too that's part of the salsa scene so we are we're not just

exclusive for one partner but just just dancing, just dancing. Yeah, exactly. Right. Well, thank you. It was a wonderful, wonderful event and looking forward to many, many years to come with my now wife, Gabi. Well, congratulations. So speaking of wonderful shows, right, this is one we've been trying to schedule for a

while. We've just had a schedule conflicts. It's another of our 101 series. And today, what are we talking about? OpenShift? OpenShift, yeah. Yes, OpenShift. What is that?

The open thing. What is the open thing? The open S. Everything is open. Yeah, open S. It's not OpenStack. Yes. And that voice.

Who is that voice? Who is that? Who is that on the other side of the ocean? Yes. Sorry. Martin? Martin, is that you? That's Martin. It's Martin Atmaya, yes. That's Martin. It's Martin Etmeyer. Yes.

Hello, Martin. Good to talk to you guys. Yeah. Thanks for the invitation. Yeah, finally. It's good that we finally made it. And I know you just came back from your little time off, and you spent a couple of weeks with friends in Poland, and then you also mentioned Croatia.

So everything… Wonderful place. Yeah, wonderful place. There's a ravine in Warsaw, you said, and then some rural places in Poland, and then you also mentioned Croatia. So everything... Wonderful place. Yeah, wonderful places. Ruin and Warsaw, you said, and then some rural places in Poland. Martin, can you quickly tell the people that are listening that may have never heard about you who you are and who you're working for, which team within Dynatrace, and then we dive into OpenShift?

Sure. So my name is Martin Edmeier. I work in the Dynatrace Innovation Lab. So what we do is we work with hot new technologies and then see how we can apply them to Dynatrace and our customer base and then scale out these new technologies into our organization and make our customers happy with it. And as you mentioned before, we want to talk about OpenShift today, which is not

to be confused with OpenStack, which happens, happened to everybody of us. So OpenShift is a container application platform, which is an open source project, just like OpenStack, but it's got a totally different use case. So OpenStack is not a pass like OpenShift or a pass like platform. It's actually on the infrastructure end of things. And you can think about OpenStack like your own, you know, dedicated AWS data center on your, you know, on-prem infrastructure, right? So you can run Open shift on open stack which is perfectly fine but not the other way around right cool and uh so that's good so it's kind of like

you know build your build your private cloud then it's open exactly but you want to build your own platform as a service where we talk more about applications or services then it is open shift yeah if you want to run your containers at scale, it would be OpenShift that these containers would run on. And when we talk about containers, is this just Docker? Or is that the synonym that we talk about here? Or is there more container technology out there that is supported by OpenShift or is it Docker?

Well, that's a great question. I mean, there's a lot going on in the container scene at the moment. Docker does not just have friends in the industry. Docker is still the leading container runtime, but it contributed to an open container format, which now gave birth to a new container runtime that is called ContainerD. And ContainerD, in the meantime, is also supported by OpenShift and the underlying Kubernetes platform. We were going to talk about Kubernetes a bit later on as well. And there's also another container runtime called Rocket.

It's written RKT, which I think has been created by CoreOS, which is a big company in the container field. But I'm not sure if that is already supported on Kubernetes. But container D, of which big parts have been actually provided by Docker themselves, by the company, is also supported container runtime. That supports the open container format. So I think this will probably change and mature over the next month even because more and more container runtimes are coming up and there seems to be a place for each of these in the industry, right?

So it's not just Docker anymore. And that's an important delineation to make docker containers are not docker but docker is container the containers are a foundation but they're and docker is a company that did a lot with them right so it's just a just a bring that point out great that's a very interesting point actually because docker was the first company they were actually creating you know docker the tool to run containers as a side project it was not the project that that they actually wanted to be successful with um but they were not successful with with the product that they've built and they've created an abstraction layer on top of what is called Linux containers.

So containers have existed for, I think, the past 15 years. But Docker as a company was the first company who made Linux containers easily, let's say, digestible or accessible by enterprise customers and by developers. So it's nothing new that they've invented, but they've reinvented how to access containers in the enterprise. And suddenly developers were able to say, well, think of consultants.

They're working on a new project, and they can simply stitch together their development environment by saying, I want MySQL, I want Java, I want Tomcat, I want Node.js. And with a click of a button, you can quickly spin up this environment and start to work on it. And then when you don't need it anymore, just tear it down by another simple click of a button, which was just, you know, a huge, huge improvement to how infrastructure and environments have been built back in these days. And I think it's already three or four years ago, right? And a lot has changed in that time.

That's true. So I was going to say, if Docker is to containers as OpenShift, let's shift now from the containers to OpenShift. Who offers it? Who makes it? And what is it? Yeah, give us the 50,000-foot view of it. Yeah.

If you can. So let me try to make the jump from Docker to OpenShift because, you know, what's important, you know, Docker has been a great tool for development to quickly define a stack, to try out containers, but it's not really a tool for using containers in production. Now, that's what OpenShift is for, or actually the underlying platform underneath of OpenShift.

So OpenShift is based on Kubernetes. And what OpenShift is, as you could put it, you could say it's an enterprise-ready version of Kubernetes, right? So let me quickly dive into what Kubernetes is about. Kubernetes has been started as a project around, I think, in 2014. It was a project backed by Google, which had more than 10 years of experience in running Linux containers at scale for their infrastructure. And they were using an internal project, which is called Borg, B-O-R-G. And Borg eventually led to Kubernetes, which was open source and GitHub, and which over time had a couple of companies contribute to the source code. And these companies were companies like Red Hat, Microsoft,

MyRentis, CoreOS, and many others, right? And so how OpenShift came along then is that OpenShift, you know, since the Red Hat, the company Red Hat, which also has a foot in the OpenStack ecosystem because they also have their enterprise-ready OpenStack offering, it stays sensed that Kubernetes might become the container platform of the future for running Docker containers. And back then, 2014, 2015, it was exclusively meant for running Docker containers

to run it in production in an enterprise ready way so again openshift is the enterprise ready version of kubernetes by adding you know security and access rules for users and groups to projects or namespaces as they are called in in kubernetes which was back then, not available in Kubernetes, right? And Kubernetes was also only meant at first to be a tool for, you know, back then they called it for orchestrating containers at scale. But in the end, it turned out to be a complete abstraction of your data center. You don't care where containers run. If you compare Kubernetes

with other solutions in the highly available Linux space, you had to assign certain weights to each of your processes and stickiness factors on which hosts they should preferably run. But with Kubernetes, the orchestration engine in Kubernetes and the scheduling engine automatically defines where this container should actually run on which hosts. So as a user of Kubernetes and OpenShift, you actually don't care about this. You just want to run a container right and so it became a complete abstraction of your data center which has tremendous you know effects on on an organization that leverages this technology because it allows for as we hopefully talk about later on for a clean separation of the duties between development and operations people right um and because you, since developers, since as a company, you don't have

to care anymore where you need to deploy a certain artifact, like a JAR file, if you deploy a Java application or a WAR file, also in a Java space, there's no need to involve any operators in, you know, placing these artifacts on the right hosts. So this leverages, this actually lifts the gifts, enables developers in your organization to take the source code, build it, and run it on this platform without any involvement of the operators, which take care of that the platform runs smoothly but you don't you know need to throw any artifacts over the wall as we did a couple of years ago

before we found out the dev offices and communication collaboration is really a good idea right so this enables and empowers developers in a way that hasn't been seen before wow um thanks for that that was a that was a cool uh i mean obviously i love that the topic devops and uh what basically you just explained you know kind of bringing depth and ops closer together breaking down these walls and you just you just perfectly explained how this all contributes to breaking down these walls because you're actually taking out the roadblocks in how you can actually get something that developers develop over on the operation side and make it smooth and easy to run.

So OpenShift, you said, obviously Red Hat, right? Red Hat is behind it. Now, people that may look into something like OpenShift, like their platform as a service, what is out there in the competitive space so that people that may have heard another name other than OpenShift know, ah, this is now where the competitive landscape looks like?

Do you have any other offerings that you can quickly highlight? Yeah, so I think that on the same level, like a direct competition of OpenShift and Kubernetes, I would totally see Docker Swarm, although I wouldn't see it as big as a competitor as Cloud Foundry. Cloud Foundry, I think you already had a 101 on Cloud Foundry a couple of weeks ago. So I will not get too deep into Cloud Foundry. But Cloud Foundry is another, let's call it PaaS platform. It's a technology which allows you to take your source code and run it in the cloud. And Cloud Foundry figures out for you which technology you're using and then instantiates, let's say, a Tomcat application container, which is not to be confused with a Docker container.

So also in the Java space, you speak about serverless containers and container engines, but that would be a process that is started somewhere in the Cloud Foundry cluster, which then runs your Java code and can even build your Java code. So that is also a very convenient way to work with technology, right? To work with your source code and to enable your development organization to quickly deploy code into production. Whereas the operations team is more for taking care about, you know, the frictionless operations of the entire platform, right?

But not so much about application deployment. So that's what OpenShift and CloudFoundry really have in common. On the other hand, OpenShift is, I would say, a more lightweight architecture. Right. I would say a more lightweight architecture. So I think you learned in the last 101 and the audience learned that it takes quite a high amount of infrastructure to run Cloud Foundry, which is partly due to the fact that a vanilla installation of Cloud Foundry, although that might have changed in the meantime, is by default in a fault-tolerant setup, which means that you have redundant components and triple modular redundancy of the important bits and pieces of this cluster, which is not something that you get out of the box of Kubernetes,

which you can do, but which is not done for you out of the box. So this has both pro and con maybe uh because you need to dig deep into the documentation of kubernetes and then openshift to get this done for you right um but this also allows you to run kubernetes and openshift on a on your laptop which is a huge enabler for companies again, because I've been talking to so many customers and prospects due to my work in the Dynatrace Innovation Lab. And what's really fancy about OpenShift and Kubernetes, if you can enable your developers to try out a new technology, and I'm talking about, you know, in the cloud-native space

where every company is a software company, you have your organization driven by your development teams. Of course, by the customers, right? But you want to make the developers happy because they create the business that you make money on. And so if they adopt Kubernetes in person and if they advocate for it, you have a much higher chance of widening the adoption of a new technology throughout the technology, throughout the entire company. And as we said before, if there is a clean separation between developers and operators, it's just another example that DevOps is not just about collaboration communication it's as Steve Nelson Smith one of the founders of the DevOps

you know movement said DevOps is about the appropriate application of both technology and attitude and this is a great example that choosing the right technology can bring so many things forward and can in the end really enable, you know, fast time to market and fostering innovation and agile teams, right? The way it should be. Yeah. And I think what's interesting is that, as you said, you can run it on your desktop. So if you want to play with OpenShift, there's nothing to stop you from downloading it, running it, and then seeing how you can incorporate that into your whole cycle there. And not to say one is better than the other, but I know even our own challenge internally is when we want to play with Cloud Foundry, we can't just spin one up, right? That takes several VMs to get going, and it becomes, you know, for, you know, for us who don't necessarily have a VM budget, we can't just say, okay, I'm going to spin up five

servers and, and deploy Cloud Foundry just to experiment. So it does make it a lot easier for developers to look into OpenShift in that, in that, in that fashion. And I think another important difference between OpenShift and Cloud Foundry, again, not opinion, not being saying which is better or anything, but, uh, Cloud Foundry is a very opinionated platform, meaning you're going to get the version of Tomcat that they're saying is the version that's going to be used. Uh, whereas I believe with OpenShift, um, you can set, set up whichever one you want. Is that correct as well? Yeah. I mean, I mean, I think Kubernetes and OpenShift

is still an opinionated platform because it mandates a certain way of how it's being used, and how efficient workflows look like. But it's not opinionated in the sense that it prescribes which technologies to use and what the platform has on offer. And I think, to be fair with Cloud Foundry, I think it allows you to choose the version of Tomcat that you want to use. But that, of course, is limited. You can't use any version. using a pure container application platform as OpenShift, you can choose any container that you would like to use and stitch together your production environment or your, let's say,

the foundation, the runtime environment of your application that you want to build as a development team based on an arbitrary set of containers.'s that's the ideal scenario on the other hand we've also learned in the past that i mean you could of course use any container that is on offer in any from any corner of the internet but you shouldn't probably do so and that's why why openshift or redhead recently came up with with a which contains, which is, you know, you just have to register on this catalog to draw, you know, the images from this catalog. But it contains a huge set of certified containers that you can use safely to build a foundation for your innovative applications. And that ranges from, you know, JBoss enterprise application servers to Nginx, even Dynatrace is now part of this catalog. And the certification means that the containers who land on this catalog are automatically checked for security

vulnerabilities frequently, multiple times per day. And when they detect a leak in your image, because it could happen that another heartbleed would appear, right? Or that a library that comes with your image that is implemented in that image that you've pushed onto that container catalog to make your application available to a bigger audience is somehow compromised and needs to be you know exchanged and so they will push you as an isv and tell you hey please martin or please you know please company x please ecmeco.com please rebuild the container based on the base images images that we provide to you and which are secured, and then your entire application will be secure again

and available for users to use in a sensible way that doesn't compromise your users in the end in your business, right? And before we move on to the next topic, I just had one last question about compatibility with an OpenShift. I imagine you could could run if you wanted to the um.net core on linux right because it's just another thing but does open shift support the microsoft operating systems or is it all linux based no no no they they do they do they have they are they just recently went into um a huge partnership with Microsoft, and I was fortunate enough to see a keynote

on the last Reddit Summit. What you can do now is you can run Windows servers as part of your OpenShift cluster, right? So some of the machines that run OpenShift would be Linux-based, and other ones would be Microsoft Windows server based right and and then when you when you as a as an end user of the OpenShift platform which would typically be a developer when you want to run a Windows application or a Windows containers a Windows container the

only thing that you would have to do is say hey please run make sure that this runs on the Windows machine and that's everything you have to do and then the scheduling and the execution of containers is again handled by by the orchestration platform right but this is something that you can do and they've showed a great demo on how to run they actually run a container live on the state on stage where they've shown a windows command prompt running on openshift that runs on a Windows server underneath OpenShift. That's cool. Which is quite impressive because, you know, a couple of years ago, maybe, you know, three years back from now,

who would have guessed that RATAT and Microsoft would be in a strong relation in a partnership that wouldn't have even been, by anyone, I think. Right. And as we saw with our one-on-one on Azure, Microsoft is making much more of a shift to getting revenue from Azure as opposed to their OSs. So yeah, quite a change there. Yeah. Before actually moving to the next topic, and I know we have a list of things you want to discuss. I have one last thing on this one. Now we mentioned Cloud Foundry and OpenShift,

but there are also other companies out there. if you look at aws right there's a lot of services out there and i believe i mean they would also claim that they provide kind of a platform as a service because they have different services that allow you to do these kind of things but maybe not in in one service but they have like all the AWS services and some of them you pick and choose. Microsoft has Azure. So I guess we should still name them as in the same kind of competitive space, right? Because if you want to build apps and you don't want to deal with standing up your physical infrastructure, but you want to just deploy your containers, then AWS or Microsoft Azure or probably even Google Cloud, I'm sure they are also to look at. Yeah, certainly.

I mean, I wouldn't see them necessarily on an eye-to-eye level from the perspective and from what they have on offer. I would see them more on the IaaS space, but they are bringing up more and more services that even leverage Kubernetes underneath to run containers. Like AWS does it with the, I don't really know what the container service is called, but on Google Compute Cloud, it's called gke which stands for google container engine interestingly it's written with the k in the middle because gce is the google compute engine already but the k in the middle also provides a hint that they're using kubernetes underneath i think this so i think that would be one of the reasons i saw a lot of tweets about that last week and i was wondering about the K.

And the other difference too is that if you're enterprise and you want everything in your own private cloud, that's where you're going to go with OpenShift or you're going to go with Cloud Foundry. I guess you could theoretically go with the Azure on-prem offering, but then you don't have control of your hardware at that point. But for those enterprise teams, it's more of a, if you want it internally, these are

more the way to go then. Yeah, and interestingly, OpenShift has a very broad spectrum of offerings. It enables you to run OpenShift in your private data center, but it also offers

commercial offerings that run pre-provisioned in the cloud for you and you just choose the availability zone that you would like them to run in and uh so they so reddit has an offer for you know basically every market that you could think of for developers trying to get your feet wet uh which would then be using opensShift Online, which is a very small kind of service, offering only a limited amount of resources. And then the next bigger one would be OpenShift Dedicated, which runs on a, I would say, considerable amount of resources, like 32 gigs of RAM per machine on an AWS, as far as I know.

And then OpenShift container platform, which would then be the on-prem version that you would maintain on your own. And I think it also supports running these offerings on Azure in the meantime. There is so much going on in the past that I think over the last couple of weeks I probably missed a couple of hot news in this regard. Yeah. But it seems to proliferate into a couple of places recently sorry yeah that's okay so to switch gears a little bit so you talked a lot about how cool it

is that we you know the devops is about you know breaking down these barriers and technology and the two the right tool selling the right tools obviously help that so now tell us and explain us quickly, if I am an organization and I want to push new services or applications out from dev all the way into ops using OpenShift, how would that look like from a very high level? Start on the development side and how does this work? So that is a great question because I think that really makes a difference

between Kubernetes, using Kubernetes and using OpenShift. So OpenShift, as I said, is the enterprise-ready version of Kubernetes. And what that that is not governed by that is not offered by kubernetes and for developers that means it means that you know operations people and the business have decided that developers can use or can choose from a predefined pool of secured and hardened images, right? So that can be based on Node.js, certain Java technologies, Golang, and so on and so forth. So as a developer, I can say, let's start a new project. Let's bring this out to market quickly.

So what I can do is I can take a look at the image catalog, which is embodied into my OpenShift cluster, which I can attach to from either the web UI, but also from the console using my command line interface as a developer. And I can say, I want to build my project on a combination of Node, Java, and MySQL. And then just say to OpenShiftift now please take the source code which i've written which is cut right over these you know github repositories or git repositories and please compile that and run it on openshift so and and i can even configure in openshift

that whenever my source code repository changes when when I make a change, please build a new image and automatically run it on OpenShift and replace any existing images in a way that it doesn't break my application, so that there is no downtime, which is called a rolling upgrade or a zero downtime release. And that is automatically handled to you by the platform. And that is just insane. I've been building such systems in a previous engagement with a different company. And we were working on this for several years. And we, of course, with many thousand developers contributing to Kubernetes and OpenShift, you could spend decades and will not reach the same level of of of you know convenience but that is just incredible as as a developer i would definitely fancy to have this convenience that hasn't been

there like three years ago right you would have to build a system yeah yeah and this is so what you just explained is it kind of sounds like it includes some of the tasks that you would normally assume a cio cd server like a jenkins is doing so listening for a git update and then triggering a build pipeline so is the build pipeline also part of openshift so it's not it's not directly a part but openshift themselves have dedicated a lot of work to a Jenkins plugin that you can just simply deploy by saying, you know, deploy Jenkins on the command line, and it will create a Jenkins installation. And you only have to configure the deployment pipeline, right? And you can even extend that to, to, you extend the continuous delivery deployment pipeline that initially comes with Jenkins to also feature tools like SonarType to do, you know, code analysis, static code analysis,

and then also, you know, leverage the OpenShift ecosystem, the partner ecosystem, and bring in, you know, binary artifacts like Artifactory from JFrog to really build an exhaustive, continuously delivered deployment pipeline with just a single click or a single command at the command line to have something up and running that companies have been struggling with to implement for the last couple of years, which is, again, amazing. That's cool. And then, so basically, it means I make make a code change the platform and the pipelines run and basically well let's say the the platform

makes sure that my code changes get deployed into an environment i assume there's a notion of there's a dev environment a test environment a staging and a prod environment so these environments are somehow separated through tags or through other means i assume yeah you would you would typically separate them through what's called a namespace which is a separation of your physical cluster into you know virtual clusters there are different ways of doing it if you if you are a smaller company i I would advise going with namespaces. If you grow bigger, I would rather choose to use a dedicated production cluster to not interfere with the compromise on the cluster performance when running too many virtual environments on the same physical cluster and then reuse one for development and pre-production as a separate entity,

a separate cluster. But there's a couple of options which you could choose from but but in the end it boils down to the same thing yeah you don't want a developer accidentally doing scale 10 i mean scale 100 when they meant scale 10 right yeah yeah exactly exactly cool and so and then eventually right i mean all whatever code i change and i run through different environments and i have my test environment and everything is good and then i just run the same pipeline basically and then just deploy it in production and if if the system and the platform ensures me that there's no impact like it's a rolling update i mean that's even more beautiful yeah and that's pretty cool and then i assume things like uh blue green deployments are obviously part and supported or i mean you

mentioned rolling updates but that's exactly how they how they implement it yeah i mean the only the only premise is that you run the container at least two times so that allows the platform to tear down a single container and replace it with the new version and once that is up and running and and serving uh the users well by checking you know health status of that container it will tear down the other ones one by one until each container has been replaced by a new version and that is all done automatically and the jenkins plugin neatly ties into OpenShift itself. So the Jenkins plugin will actually, you know, trigger those changes on OpenShift when, you know, all the stages which you defined as your, you know, gates to production have been passed successfully. Cool.

Yeah. Martin, I remember, was it maybe uh when you were at the open shift summit in boston the reddit summit yeah yeah reddit summit yeah so i remember you gave a talk uh because i want to i want to quickly touch an area obviously that is very interesting for our listeners you gave a talk which i thought was fascinating because you walked through different stages on how you can do all sorts of monitoring, log analytics, but monitoring of different components. And you can do it all with different sets of tools. But then obviously you say, well, this is – it's all doable. But it just means that if you allow your developers to choose all these different technologies, right?

Somebody may use MySQL. Some uses Java. Some Node.js. And then you have different tools that you can all – of course, they're all there. But it just means whoever takes care of monitoring and log analytics, you need to figure out a way how to do all this and how to make this seamless and to make a sense of all this different data that comes in from different technologies. And then in the end, obviously, modern APM tools like Dynatrace just solve that exact problem of providing monitoring in a very technology-diverse environment, which we have now with the flexibility that the platform like open shift gives us right it's great that we have diversity but diversity also means

more complexity and i thought you did a fantastic job in that session i think it was recorded and the slides are out there um and this is just a uh the slides yeah, on your slide share, I assume, right? Exactly. It's called Managing Logging and Tracing on OpenShift. Yeah. So for me, this was one of the most eye-opening things I've seen at this conference, especially the way you kind of walked people through it. You can do all this with a lot of open source tools, but there's a reason why professional commercial vendors like us are out there. Now, from a monitoring perspective, just briefly, what are the things people need to think about when they are thinking about OpenShift and defining the monitoring and logging strategy?

Yeah. strategy yeah yeah i think this is one of the really one of the hardest problems these days and and the industry is now really understanding as the adoption of these past platforms is increasing and really you know gaining traction in the enterprise who monitors all this you know and uh and although you know we have a lot of monitoring open source solutions governed by the cloud native computing foundation which also governs Kubernetes and which is a partner of Red Hat and working closely together on the vision and so on. They are still providing tools which give only insights and vision into separate layers of the cluster. So what's still missing is a complete understanding of how these things fit together. How does a log message tie into an issue that happens at eight o'clock in the

morning to a certain user with that username, right? For example, those are things that go through the entire stack and we have good coverage in the open source market on the horizontal level, across nodes, for example, across processes, and maybe also across the application, but not in the other way around, like from top to bottom and bottom up again. enormous efforts into creating a market-leading solution that really does that by simply providing a single artifact that you can deploy in an OpenShift cluster and that will, you know, instrument all the important bits and pieces and cover those real-time dynamics that we talked about, like scaling things up and down, replacing containers with new versions. Which are those things that you want to have covered? Because in the end, you know, in a microservice environment, and that's what these platforms are about, they shift complexity from the application into the infrastructure, you know,

by, you know, communicating, by services communicating with each other, by scheduling containers at any point in time. Things could break all the time. So you have to be prepared for things to go awkward at any time. And if you're not covered, you will never be able to find out what actually happened. Because, you know, you will end up with millions and trillions or whatever lines of log messages that will be hard to interpret if you don't have the right solution in place if you only focus on on those horizontal layers of action and not on what's what's really gluing them together and how does how do these things interact with each other's hey i think you mentioned it but maybe

for the audience to to repeat so in order to get – and I know we try to make this podcast not about Dynatrace, but obviously we have invested a lot into this and solving real-life problems. So in order to get monitoring from Dynatrace into your OpenShift installation, can you remind me again what you need to do? Where would you need to install the one agent? Yeah. So I would recommend looking up the Reddit container catalog, which you will find on access.reddit.com. And if you search for Dynatrace, you can find the certified image, container image for the Dynatrace one agent. Now, the Dynatrace One Agent is a technology provided by Dynatrace that can be

installed as a container on your OpenShift cluster next to your application containers. But by means of configuration, we can make sure that the One Agent technology can proliferate or install itself onto each node of the OpenShift cluster and then provide the data from the data center that your hosts run on, from your hosts, from your processes, from your services, from your applications, into a single pane of glass that is provided by the Dynatrace UI that will enable you to get a real-time model of the entire infrastructure, including the applications running on top, and also

including the user actions that your users are performing. And it also allows you to identify problems as they occur in any place in your cluster. is running an open shift, how they tie into the big picture and how errors and failures, how they make it through the cluster and how they affect your use and how they affect your infrastructure, which is just amazing to see. Most people that I talk to, when I present them how we can do a real-time problem analysis that really helps you to dig down into the root cause of a problem it's not an eye opener it's actually you know like a mouth opener

they they stand there with with a with their mouth open and say this is the best thing that i've ever seen that's cool um hey martin um yeah i want to make sure that people definitely know where to look for more information. So we mentioned on your slide share, and we put the link up there as well on the podcast. This is where I really encourage everyone to look at your how to do logging and monitoring in OpenShift. This is a phenomenal one. Any other information that you want to make sure people look at when it comes to either any information around OpenShift or in particular monitoring? Any other hints? So if I were to give only a single recommendation, I would certainly recommend to look up dynatrace.com slash openshift, which will lead you to a landing page that gives you more information on what we have on offer.

And we'll also provide a couple of links for you to learn, to dig deeper into what we have on offer. And it will also allow you to try out our product. And we also have a couple of great blog posts on blog.openshift.com. Just look for blogs from the Dynatrace team. And we have another great resource on blog.dynatrace.com. If you look up OpenShift monitoring with someone from Red Hat who's leading the partner ecosystem as a technical director.

His name is Chris Morgan. And together we had an interview on the ins and outs of running microservices at scale and how OpenShift and Dynatrace are a good fit to solve the problem of how to become more innovative and achieve faster time to market with these solutions together. Cool. Awesome. Awesome. Love it. Andy.

Ryan. I know we tried to do the 30 minutes, but we didn't. You always say you always have to mention 30 minutes. So that's the running joke. It's the running joke. This episode was 30 minutes. Well, you know what? If people play it at like two or three times speed, which you can do on podcasts, they can bring it down.

I know. We'll talk really fast like this. How are you doing? All right. I know. So anyway, I think it's time for the Summaryator to make his appearance. Let's do it. All right, then.

Here we go. Okay. Well, first of all, this was finally another eye-opener. Thanks for the introduction on what is OpenShift really all about. Also in the beginning, making sure people understand OpenShift is not OpenStack. We're talking about two different layers of your stack here, to be honest with you. We also heard about the other players in the market, right? They talked a lot about Cloud Foundry, but also the other players like Amazon, Microsoft, and Google.

They try to provide services to make it more convenient to solve one big problem, which I believe Martin explained very well, which is OpenShift really helps DevOps organizations to break down or get rid of roadblocks between development and operations, making it easy for developers to focus on developing code on a set of pre-selectable containers, right? Those that are in your repository. So they can select from a set of containers of technologies, develop the code, and then the code automatically gets rolled and pushed out into the platform, whether it's a development, a testing, or the production environment, without interruption of your operations. And I think this is a nice place without operations having to learn more about what these technologies actually do and how to deploy a jar file. Everything is containerized. I think, Martin, you had one great phrase in there. You said OpenShift provides a very nice user experience for developers to push their code changes into production.

And I think this is a great way of explaining it. I think you also gave a great overview of that OpenShift is enterprise-ready Kubernetes and gave a summary where Kubernetes came from. implement DevOps. Implementing DevOps might be the wrong thing, but making DevOps a reality, a better reality in most organizations by reducing the time it takes from the inception of an idea until it's out in production. Well said. Very nice.

So my only thing I'll contribute, thank you Andy for another amazing summary. The only thing I contribute is, thank you, Andy, for another amazing summary. The only thing I contribute is, you know, it's amazing that there are so many options out there right now for these kind of deployments. One of the obviously one of the nice things about OpenShift is that it's very easy to run on your own laptop and play around with and check it out. I know a lot of the other offerings have free trials that you can go in on and play around and just, you know, from a development point of view, get used to the idea of this whole, how do we push and deploy bills and all that. But what it really ends up coming down to then is, you know, all these offerings have some slight differences. I think there's a lot of them are starting to blend in their offerings, right? As you mentioned earlier, Martin, there's the

Microsoft capabilities of OpenShift now, which I think a while ago they didn't have. So that was one of the differentiators, but now they're up to par on that side. It sounds like it really comes down to finding out what that full feature set is, how are you going to plug it into your pipeline? How are you going to do, you know, operations is going to leverage it, you want to run it and then figure out which of these past type of offerings that you're going to, you know, what's going to fit best within the organization. But it's exciting that there are so many options out there currently, and they're all really pushing really, really hard to really just own everything, which is really cool. And then obviously there's the other, the more important thing coming down to the monitoring is right. There's always the idea of monitoring your applications.

But what people have to remember when you're running something like OpenShift is you have an infrastructure and you have to not forget about your, your, your, your paths infrastructure, how that's performing as well, you know, and, and, and you want to make sure that some sort of uh that no component of that open shift backbone is causing an impact on your applications that you're trying to run or vice versa let's say you do have a problem on your open shift platform knowing whether or not that's actually impacting your applications or if it's not is very important. So there's kind of two sides of monitoring there. There's the open shift administration team who needs to monitor that ecosystem.

And then there's the entire business who has to, you know, monitor the application side of everything. So it's, it's, you know, monitoring is always very important in all of these whatever technology you're using just don't forget to monitor. And that's all I have to say. Martin, Marin, any last words from you, Marin? No, but it was awesome for having me. It was great to be part

of this podcast today. I'm very happy to be with you. And I finally made it. Excellent. Well, thank you for taking the time. Anything else comes up in the future, Martin? You're always welcome back. Any new topics? Any new stories? let us know. Cool.

Thank you, guys. Thank you very much. All right. Bye-bye. Bye. Bye.