Cloud Native Computing Foundation KubeCon + CloudNativeCon North America 2018 (Seattle), 16 Dec 2018

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Intro: CNCF Trail Map - Dee Kumar, Cloud Native Computing Foundation

Description

Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io

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

Intro: CNCF Trail Map - Dee Kumar, Cloud Native Computing Foundation

The interactive landscape -- l.cncf.io -- allows you to filter and sort hundreds of cloud-native projects and products, and review details including GitHub stars, funding or market cap, first and last commits, contributor counts, headquarters location, and recent tweets. Engage in a small group to discuss how the landscape was built and what should change about it going forward

To Learn More: https://sched.co/Grcf

A

So I'm with the cloud native computing foundation and I run marketing there prior to joining CN CF I used to work at docker and before that I used to work at other enterprise. Computing companies like Equinix, HP and Lenovo so wanted to welcome you all and hope you guys are enjoying the sessions here at cube, con cloud native con and Seattle. So today, what I wanted to do is to give you all a brief intro to some of the tools that we have created within the cloud native computing foundation.

A

So the cloud native landscape can be pretty complex and very challenging, and this kind of is an eye chart and that can explains how complex or challenging the overall cloud native landscape can be like. So what we did at CN CF, the cloud native computing foundation, is we created a resource map which essentially represents all cloud native technologies, products offerings, including vendors or members, or major cloud providers, and their products and services in a very interactive way for anyone, who's new to cloud native can access this resource or tool.

A

So we are really proud that, because there's a lot of work that went into it, we started this resource map a year ago, and originally it was just a PowerPoint slide.

A

We made long and made this very interactive in the sense that, if you want to update a specific category- or if you want to update a specific product or service offering all you have to do is to put a request on a pull request on github and we'll make sure that all these shoes that anything that you see here will be corrected. The other fascinating thing about and I want to just launch the interactive landscape. For you also.

A

So, as you can see here, so the URL is landscape, dot, C and C F dot, IO or you can just do L, dot, C and C after nioh, and here there are many different filtering options and it's pretty fascinating when I kind of just went through all the different options. Is it pretty much shows and let's just dive into one of this, which is envoy, which was just the project that graduated last week as you can see, what we have done is we pull information from various different sources here?

A

So, as you can see, it tells you the first comment. The last comment: the latest tweets around this project and if it is privately funded, it gives you more information in terms of the amount of funding raised or if it's a public, offering we kind of then talk about the overall market cap associated with the specific project or technology. So pretty much, it's one single view and you get everything you need in terms of understanding or coming up to speed with cloud native technologies or projects.

A

The other thing that we have done is you can see there on the left that there are many different filtering options here. So you can sort this by either a cloud service provider and if you want to look at all their products or service offerings and then what we have also done, which I'm going to kind of take you all through is essentially what we call like the overall or a well-travelled or a well-tested path in the overall cloud native journey.

A

And then there are different steps and what we have done is we've done a detail view into each of those steps. So you can, if you're, looking for a monitoring solution and when you click on it, you will see a bunch of monitoring products that become available or, if you're, looking at run-time contain I'm solutions. I've just chosen that it automatically updated the view, and you can see that it shows CNI rocket and container D. So here are just some examples in terms of the filtering options that are available.

A

So just this view in terms of the overall representation in terms of how big the cloud native market cap is it's about in terms of it, shows it's about a trillion in terms of the overall market, cap and I.

A

Think if you just look at all the privately funded open-source projects, it adds up to about 13 billion dollars, so would HIGHLY encourage that you guys can go around and play with the cloud native, interactive landscape and again, if you see any products or services or categories that need to be adjusted or need to be newly added, please feel free to go to github and submit a pull request. There.

A

The other thing I also wanted to share with you all is so one is the overall cloud native interactive landscape.

A

We just talks about all the open-source projects and technologies and then a subset of that is around serverless, so we created a service working group and as a part of that, we have made available a service landscape, and so you can again see this available when you just choose or filter out the surveillance options as available as a part of this landscape and again that gives you just a view into surveillance and again we have made that available as a PDF.

A

Access to the front page, which is the interactive landscape and then the back page, is the trail map, which is essentially what I'll be again are taking you all through.

A

So one of the things we do at CN CF as an organization is we work with the entire community and the community developers, contributors, maintainer, x' project leaders, program managers and we talk to a lot of enterprises. We work with all the leading members and the cloud service provider.

A

Think, though the cloud native landscape is still it's I would say it's still maturing and most of the large enterprises to find the journey to cloud native or the journey to digital transformation, pretty daunting. So as a first step, one of the tools that we have created as the cloud at the cloud native computing foundation as the strain map and again this is one of the handouts that I was referring to, which is available at our CN CF booth.

A

So the trail map is again a and it is a recommended or a well-travelled well-tested path that we think there would be a good starting point and again there are different paths, different use cases that one can take as you get on the cloud native journey. This is more like a baseline or an initial starting point and again I.

A

Don't think I've seen any enterprise company that is using all the projects, so we have about 31 CN CF projects hosted under CN CF, but I don't think all thirty-one projects can get used at one Enterprise company. It's going to take multiple years to get there, but at least most companies are still I would say the first three steps, which is containerization CI CD, followed by orchestration I'm, going to just take you all through the ten steps that we have defined in the cloud native trail map, which is essentially a starting point.

A

So, as you can see the step, one is containerization. So as well as all of you know, you can do cloud native without continued rising your application, so any size application will do so, even if you have a PDP 11 code, which is running in emulator, which is what Ticketmaster actually had and they took that whole code and they put it inside a container and that's how they're self-serving billions of users were actually using Ticketmaster. So the first step is to containerize your application.

A

The second step again, there was a bit of a debate in terms of do you set up CI CD, or do you once you put your application inside a container, then do you think about orchestrating it so turns out again having received the feedback from the community and also looking at various use cases and talking to several customers. What we've learnt is I think the best way to set this up is one once you containerize.

A

The second step is you want to set up your CI CD workflow, so you want to make sure that all the changes you do to your source code automatically results in a container getting well tested and deployed across various environments, whether it's staging QA or production. The second thing you perhaps want to do is to set up automated roll backs or rollouts. So today, at CN CF, we don't offer a.

A

Definitely there are other solutions to look for which is from Jenkins gait lab SIA in spinnaker, and the third step is orchestration. So once you have your application containerized, you set up your csa d workflow and the third step to look at is orchestration and definitely kubernetes is the leading orchestration platform. I just pulled out some recent stats today. So kubernetes has hundred thousand comments, 85,000 pull requests and about twenty five thousand contributors or developers and kubernetes stands number nine for comments and number two for authors or issues. Second, only to linux come.

A

That existed eight, so once you have evaluated other orchestration solutions that do exist, which is me so slouched, Nomad, Dockers form or kubernetes. It's entirely up to this specific solution that you're looking for is to start orchestrating your clusters or your kubernetes clusters.

A

The step four is observability and what we kind of represent this in to simplify the language around this, because three other types of solutions that fall under observability. It's essentially table stakes for cloud native at scale, so some of the solutions under this category that we recommend that you look at or consider is Prometheus for monitoring again Prometheus was a was the second ciencia project to graduate this year. The first project was kubernetes.

A

Prometheus is a monitoring solution for cloud native applications and once you finish orchestration and you're thinking about monitoring the next thing to look at is logging, because, once you have all the data collected, flew Andy provides a unifying solution for collecting all types of data and then glue. Andy is an open-source. Fluent e is an open source cloud native project and then the last step is about is around tracing.

A

So what we recommend is to look for a open tracing, compatible type of implementation like Yaeger, so the name Jaeger means hunter and Yeager cave was a open source project that was built and tested from uber and so eager is now a very popular open, tracing type of solution to look at and the other thing I wanted to share about a Jaeger the way they were kind of looked at the open tracing or the overall open, tracing or cross distributed systems. Is they used dapper, which was a google internal tracing system as an inspiration?

A

And then the other solution that exists today, which is also a very popular open, open facing project as which is built and tested by Twitter? Is that.

A

So the step five is now: you start getting into a little bit of more complexity, which is all around service mesh, which is about being able to connect various services together. So you can constantly monitor ingress from the internet, and this includes health checking or routing or load balancing between multiple services and some of the CN CF projects that we recommend that you look at here is envoy again super excited about envoy, which graduated last week and Matt Klein from lift.

A

He has done a phenomenal job with just talking about several different use cases with envoy, as well as he's done several keynotes now and talks and there's also a blog post, which kind of summarizes the use case.

A

The multiple use cases with envoy, so the other so essentially envoy, is a modern edge and a service and a service proxy, which is there was a design for cloud native applications and then the other thing to look at is linker D, which is a network proxy like service, and it is designed and deployed as a service mesh, and then it helps to control services inside an application.

A

So the step six is networking, so in order to so networking can get pretty complex. So there is. There are three open source projects to help build that next step as keep moving up the stacks when you start from the info level and you keep moving up, is to consider looking at a flex, flexible, flexible networking layer, usually most of the solutions that we recommend are CNI compliant. So that's you could look at calico, flannel or weave net and then you look at step. Seven, it's all about storage and distributed databases.

A

So what this is not there on the slide, but rope is a cloud native storage Orchestrator for kubernetes and Vitus gives you a lot more options than a single database. Well, so, if you're, primarily looking for more resiliency or scalability, an option to look at is Vytas and it's a really good option for running mice, sequel at scale through sharding.

A

And now you get into the messaging layer where you do want to have all these different micro services talking to one another and across multiple different environments and even platforms, if you will so, if you need a higher performance than json rest, you should consider using g RPC. So again, g RPC is a high-performance RPC framework which enables communications between your library's clients and servers running across multiple platforms.

A

The other messaging system is Nats Nats as a CN CF project, and this is a publish and subscribe messaging system, primarily oriented for the metal layer and Nats a messaging system. They have several different use cases and one of the prime use case that I've seen is around Internet of Things.

A

Then step nine is container runtimes I. Think most of the solutions that we have listed here are something that you guys must be already familiar with. So, if you're looking for an alternative container runtime, the thing to consider is to make sure they're, OCI compliant, and some of the examples here are container D rocket at cRIO.

A

Alibaba's pouch container runtime is also another good option to look at so the last step is around software distribution. So again we're when we are talking here about software distribution, it's primarily for a very large scale, type of deployment in production, and so, if you need to do more of a secure software distribution, one thing to look at is notary. So it is an implementation of the update framework which we call it as tough.

A

So tough is again designed for enter paper, enterprise-grade, secure software distribution and the main job of tough or even notaries, to make sure that, as you pull the images from darker it's to make sure that it ensures that the image is signed, its trusted its verified, its untampered as the application goes through multiple stages during the CI CD workflow.

A

So and tough again, it's like a set is designed for enterprise-grade and the way it does it is. It uses various cryptographic keys to make sure that the signing and the content verification is fully trusted and signed and untampered, and we are seeing deployments of tough, that's being already used in production like like docker VMware leap and at the other companies that are listed here and just recently. Harbor is a secure container registry and this was an open source project from VMware China, and this project was now recently donated to the cloud native computing foundation.

A

So some of the enterprise use cases that we are seeing is companies would like to offer a harbor a secure registry in order to address certain regional compliance, type of requirements.

A

So, as I said so, the cloud native trade map summary so BIA gives you this as a handout. So this is just a very brief intro, but this is a really good starting point, because I know that things are quickly changing in the cloud native landscape and it can get very complex and not knowing where to start. So this is a well-traveled path again, I get asked all the time like. Is this what we need to be sharing with our enterprise customers, and is this the only path that they need to be taking?

A

Absolutely not so there are multiple different paths to take. So one of the things that we are also looking to do the cloud native computing foundation is to create a next version of the trail maps. I would be happy to take feedback comments or any suggestions you might have. So a couple of things to be thinking about is to make this interact. The trail map not just a slide, but something that's more interactive.

A

The other thing that we're also looking to do is to really bring out like the top and use cases and also top ten problems or challenges that enterprises are facing today and I. Think now we are in a good position to further evolve the cloud native trail map, because I think now one of the recent surveys that came out from red monk.

A

It said that about 50 percent of Fortune 100 companies are using kubernetes in production at a massive scale, so I think now we are at a point that people are getting still are familiar with at least the first three steps, containers, CI, CD and orchestration, and definitely after that I think there is a lot of permutation combinations that can occur and, like I said, you'd necessarily don't have to use all the different projects to get to solve a specific problem. So we are looking to make this very interactive.

A

We are looking to actually make it in a way where you can look at a specific problem and then look at. Maybe it should I consider this project and what would be a potential outcome.

A

So this is going to be a really intensive, complex project because we'll be needing a lot of input and feedback from the community, and we are also looking to talk to several enterprise customers and the other thing that one of the communities that's very close to CN CF is our end user community. So you might have heard this through our various sessions here as well or you'll, be hearing it in the sessions you go to today, so end-users. The way we define them are they are users of cloud native technology.

A

So that's the main difference between an end-user versus a vendor, a vendor or a member could be somebody like the top providers. It could be Amazon Google, where they're, actually commercializing and selling products or services or solutions and the main difference is the end users. They don't actually productize or commercialize and sell, but rather they're primarily using cloud native technologies internally and they're, trying to help optimize their all your internal operations, so whether it is from a CSE D perspective or whether they're looking at putting together a brand new compute platform.

A

So we are looking at a lot of end-user use cases and they're trying to see how best we can leverage the learnings from the end-user use cases and eventually applying it to the trail map, and that was a similar approach.

A

We took for the overall cloud native landscape, so it would be more than happy if any of you had suggestions or any specific thing that you'd like to see as a part of the interactive version and again, if you have any questions or even suggestions, but the existing static cloud native trade map feel free to reach out to me and I can share my Twitter handle or you can also reach out to me by email.

A

So, lastly, the last tool I wanted to share is called the dev stats.

A

So this is a tool that we created, I would say three years ago and it's slowly evolved and I would say. This is an amazing tool for everyone to actually go and try it out oftentimes. We always get asked in terms of because it does if people want to understand how what is the community engagement with these projects? Are these projects healthy? So in order to do that, we have created this dashboard and again this was a very complex project that we worked on and it looks at github stars.

A

It looks at github comments, and so it's essentially a dashboard, so I'll just take you through just kubernetes for now. So, as you can see, it kind of gives you the community engagement and sizing. So you can also look at how many companies are actually contributing to kubernetes, so you can look at an hourly view or you can look at weekly.

A

So there are a lot of options in terms of how you can you can actually create your own dashboards and really understand the usage of these different projects and the other thing that I thought was really interesting. Was you can also sort it by github stars in there, in the repos and in terms of other stars are interested?

A

This kind of shows you how complex or how the usage of kubernetes is, and you can also actually Traverse back in time from the time kubernetes was an incubation project to now, which is a graduated project, and you can actually break it down by the number of comments.

A

You can also look at some of the top companies that are actually using or actively contributing to kubernetes, so I often get so when people come to me asking for the health assessment or the community assessment, and how do you actually measure if these projects are really doing well or they slowly tapering off or are they go? Is there or is there like a slow time for the some of these projects? I would say this is a public URL.

A

Anyone has access to it and feel free to kind of look through and filter it out and use your own filter criteria or sort a sorting criteria generate your own dashboards and again, if it's to be also kind of publish our monthly or quarterly reports around the projects. But we again use this as our main source. So I really wanted to make sure that most of you are aware that this tool exists. If you want to pull out your own reports, the other.

A

So this is where I got my start in terms of the number of contributors that are actually contributing to kubernetes.

A

So again, like I said, there are multiple different ways in terms of you: can slice and dice your criteria and if you want to start building your own reports and looking and gauging in terms of how all these projects are actually performing.

A

Guess that's all I had so I just kind of wanted to quickly summarize, so one is the overall landscape that we have created. I think it's an amazing resource to get started with in terms of just getting a sense of all the players in the cloud native landscape, various product services and offerings.

A

The second one is just a trail map, which is your journey to digital transformation that support enterprises are looking for, and this is just a very basic or a starting point and a static view into the ten steps in the trail map and then the third our tool to look at is what we call the dev stats. So it was primarily we started off to show the project and the health around kubernetes. But now the dev start stool has evolved to support all the projects under the CN CF umbrella.

A

Happy to take any questions, our feedback, with suggestions that you have in terms of the three resources that I shared with you all today,.

A

Yes now so it was in-house, but again the Oh.

A

So the question as I. Thank you. That was a great question. So the question is when we created when we builded appstats tool, was that done in-house? Or did we also take a lot of help with the overall community engaging and building? So this was done in-house, but then it did take input from the community, but I think the main resource that we looked at was also github.

A

So we considered we looked at all the repos and then we looked at all the stars, the commits, and then we built a way for it to automatically pull reports or data from github and again. We also, then overlay that with a lot of community-based efforts based on the working groups and some of the contributor experience groups and then so I think I would say it was like a three step process. But the overall implementation is in-house, and it's done by a guy by named Lucas many stove at CN CF.

A

And so, if we have have any changes to Death Star's Lukas make sure that again you can submit a pull request and those requests are taken and handled right away.

A

All right again so feel free to reach out to me, my twitter handle is DS printer and any questions or suggestions you have with respect to the resources that the CNC f has built around cloud native would be greatly appreciated. Thank you.