Cloud Native Computing Foundation Kubernetes Forum Bengaluru 2020, 27 Feb 2020

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Sponsored Keynote: Kubernetes Deployment Made Easy for Complex Applications - Girish Dhanakshirur

Description

Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects - Learn more at https://kubecon.io

Sponsored Keynote: Kubernetes Deployment Made Easy for Complex Applications - Girish Dhanakshirur, Distinguished Engineer & CTO | Cloud Pak System, IBM Mobile Foundation, IBM Hybrid Cloud

Containers are the new fabric for cloud native applications, bringing simplicity in packaging and portability to developers. However, container images by themselves do not address key enterprise needs for production applications. Beyond containers and Kubernetes, developers need to orchestrate their production topology, and to provide management, security and governance of their applications. Girish Dhanakshirur will share the lessons learned during IBM's journey of containerization and how IBM sees Kubernetes fit in the broader IBM strategy.

https://sched.co/YW43

A

Good morning, everyone for the next, maybe about five to six minutes. You know I'd spend some time so by the way I'm one of the CEOs in the cloud division, where the week spent good amount of time playing with some of these technologies building. On top of this and again, you know contributing back back to the community so for the next, maybe five six minutes. You know I'd spend some time on applications, application platforms and you know how that that fits into into communities.

A

I do participate in many of the meetups here in Bangalore I'm based out of Bangalore. So you can hit me on Twitter, Gary's, deed, tweets,.

A

So I mean, like myself, I think many of you have been. You know: building applications for quite some time, I'm guessing so so, perhaps some 10 to 15 years ago we pretty much had you know just two technologies. We were building applications right, so there was Java or there was dotnet right. So those were the two technologies or two two languages. Where we were, you know, building applications the developers were focused on building applications and the heavy lifting of deploying those applications was done by the underlying application platforms.

A

So the role of application platform was doing. Essentially, you know heavy lifting from security, scalability integration and in doing most of the boring job. You know, while giving the time back to you as application developers to focus on you know, building best of the best applications. So what has changed? You know over the period of last. You know maybe about 10 years today we are doing the same thing or somewhat similar in the cloud in the cloud era right.

A

So if you're building a similar platform to run cloud native applications, you know so that again you know same things around security. Scalability, you know are available for these cloud native applications, so there are some expectations from from the application platform. So DIMMs talk talked about. You know what makes the kubernetes and you know what what it forms is that as a base. So there are certain things for developers where you know developers ask for, you know, can I build a stateful application? Let's say a database can I build a web application.

A

A stateless web application or you know, I- want to build an event-driven application. Let's say in a service application and of course you know the platform has to be. You know very simple and flexible. So there are many application platform vendors. You know focused on building, you know all and and addressing all of these requirements, so in cloud native world. So we see in essentially that these three styles or these three programming styles right so there's one which is twelve factor applications. You know which is primarily driven by cloud foundry.

A

You know the project in CNC F, then there's cloud functions. You know from IBM as your functions from Microsoft lambda, so these are service applications and the last one is about container izing and deploying container based applications. Of course you know on top of kubernetes, so the good news is, you know over the period of let's say last three: four years the community has come together, you know, while there were many projects for doing many things, the community has come together in order to create you know some sort of you know, opinionated stack.

A

So, for example, you know we were more or less settled on container engine to be container d and again you know, we're sort of you know settled on the orchestration layer based on kubernetes and then, of course, you know there are new projects like service mash service, catalog came native, you know there we are converging on. You know many of these projects coming together to create that opinionated stack. So you know we're proud to say that IBM has built.

A

You know that particular stack and today, IBM container kubernetes container service on public cloud is built on the stack and we want millions of applications or billions of containers. Today, it's not just running your applications, so we also have built in a good amount of our investment. You know things around Watson things around. You know: machine learning, services, things around databases, Cloud Foundry functions sto.

A

Many of these today run on the same kubernetes platform, and you know I don't want to encourage most of you in the crowd to also participate, and you know, contribute towards this upstream projects and actively I mean just like you know what we're doing at IBM. You know I would look forward. You know for many of you to be also actively participating in these projects. So let me switch the gear a little bit.

A

You know, of course, that's our public cloud story and you know share a little bit about what we do with our clients when we look at our clients again I'm bringing this up, because you know many of us here in Bangalore. You know working in most of the GSIS and there are many of the other captive companies. So we focus on this. You know this type of projects, so there's a customers are looking to move.

A

Existing applications over to cloud continue, rising them in or building brand new applications using cloud native technologies and run it on containers and the last one is about. You know how do I manage this massive infrastructure? You know running very complex applications, so I'll illustrate this with. You know a very simple example: you know what we did with American Airlines.

A

So if you look at, if you, if you look at this picture, it's about using what we call as a strangler pattern, where you run both new and existing in parallel- and you know, while you know they continue to coexist, you know you slowly turn off the traffic coming to the old and then you know you send it more and more towards. You know the new applications while delivering features to both until at a tipping point where the the traffic you know you have.

A

You have a critical mass of users using the new, and then you can turn off the old. The reason I'm bringing this up is because today, what we have in the community, what we're building around kubernetes- and you know many of the projects- many of the technologies allows us to do many of these or these patterns, so at IBM. In addition to what we do on public cloud, viewing heads down for last couple of years, container izing our entire middleware stack to be certified on kubernetes right.

A

So this, essentially you know we're taking our existing products and products, portfolio and container izing them and what we call these as cloud packs.

A

So we have about six of them today, so these cloud packs basically range in all the way from delivering offerings from Watson to WebSphere, to Watson business process, automation to block chains to api's to mobile services, so we embarked in a whole slew of these things, we're building and we re basing most of these products and middleware on kubernetes and the fact that we are running this on top of this open source platform, you know allows us to scale very well and you know, participate in broader ecosystem, so these cloud packs are certified secured and you know ready to go on day.

A

One- and you know these cloud packs are available for customers to run it on public cloud, private cloud, on-premises or whatever. So so, we've been very active, you know in the community, and you know we continue to work with CN CF and see many of the new projects or upcoming projects and actively participate, and you know we look forward to you know working more more closely in the community. Thank you.

A