Cloud Native Computing Foundation CNCF Webinars, 2 Mar 2017

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Webinar: What is Cloud Native and Why Should I Care?

Description

https://www.cncf.io/event/webinar-whatiscloudnative

This talk will provide insights into the history of the foundation, its structure, goals and the processes for accepting new projects and new members.

The Cloud Native Computing Foundation was created with the charter to define the term cloud native and provide a home for the cloud native, open source projects such as Kubernetes, Prometheus and others as well as to push the entire software industry to utilize modern cloud technologies more effectively.
Join us for KubeCon + CloudNativeCon in Barcelona May 20 - 23, Shanghai June 24 - 26, and San Diego November 18 - 21! Learn more at https://kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy and all of the other CNCF-hosted projects.

A

This is the fourth gladinet cloud native computing foundation, webinar, as you will have seen if you're looking at the chat, I've put a link in there to the previous webinar, so I would encourage you to check those out and also I've put a link in there to the next webinar that we're gonna have after this one, which is with William from the newest project that we've announced as the seniors product, the CN CF and that's link at D I suggest that you go and sign up for that right now.

A

Today, however, we have Alexis Richardson, who was a chairperson of the cloud negative computing foundations, technical Oversight, Committee he's also the CEO of weave works, he's going to be explaining what cloud native is and why you really should care about that in previous webinars. If you've been here before you will have, you will have seen that we normally ask you to keep all questions to the end, but I've changed my mind on that one, because I thought that it meant it made it a little bit stale.

A

So during the webinar please use either the Q&A box down the bottom or the chat window that you've seen I've already been typing into and I will find opportune moments to interrupt Alexis and ask him questions. We're gonna go ahead and get started. Some people are still joining but I think it's I think it's better to just kick it off now. Alexis. Are you ready hi mark? Can you hear me I can hear you just fine ready to go. I am.

B

Ready to go excellent, take it away! Thank you, Mark! So hi everybody, my name is a Lexus I'm, the chair of the TOC for CN CF, as well as the CEO and co-founder of weworks I, see that the date is wrong on. There should say twenty second of every day, oh and this correct twenty-third there you go just to let you know, if you do, you send questions on the chat. Room, I won't see them family in a full screen here.

B

Display of the slides, so mark will see them in a pass them through to me on a slack window. We have open separately, if there's a delay, don't panic. So what is the cloud negative competing foundation? It is all about open source cloud competing for applications, not the same thing as OpenStack, which is about infrastructure, for example, and our goal is to curate and promote a trusted toolkit for modern architectures.

B

So what is cloud native? This is the concept pioneered by several people around the same time that have 5 to 10 years ago, including Netflix, which I think is probably the largest and most famous early company to do this, and for them it was a very practical matter. They were going to build a global platform for delivering video to consumers, ie us that would that just worked, which meant it had to be always up and they had to keep on improving it.

B

So the customer experience got trim better and better they've talked about this many times. I particularly recommend the slides, linked here from an Amazon event to looking about what they consider to be cloud native at the time web scale global highly available consumer facing which brings with it particular needs and the key. The key to all of this is they wanted to achieve two things at once and do so at scale, and those were speed and access, meaning specifically, they wanted to deliver.

B

Change is for their software much faster than anyone had done before up to thousands of times a day in order to make for better consumer experiences, and they wanted that.

B

The accessibility of the system to be very high, so they were not willing to tolerate outages because they really felt the just the CEO in a letter to investors once described it as the moment to truth, when you plop down on the couch and you're thinking, you know what am I going to do now, shall I read a book or shall I watch something on Netflix and if it just works, that's very appealing to you as a potential Netflix customer, and they really thought it was critical to being a good provider of video.

B

I know, of course, they've been very successful in that model and it's led to other people copying the approach of being always on and very rapidly iterating their systems using the technologies that we'll talk about. And do you read these slides if you want more detail about how they see the story, so this concept of cloud native moon brew and the sort of techniques really power a whole industry today, very much headquartered around islands of technology like Silicon, Valley and places in China or Europe, etc.

B

But you know growing all the time and what's happening with with the cloud native foundation to, and are we talking about this really democratizing these tools? Does that mean, of course, that anybody can be like Netflix? Well we'd like to think so? Even these guides we can dream of not being in dunkey anymore, but in fact all being sparkly unicorns, but in reality there is this need for speed in that I rough reference before and how does that play out? Here's some numbers from the puppet lab survey state of DevOps 2015.

B

They do this once a year. This is this is a good snapshot. I took recently comparing what what they consider to be low and high performers in important metrics and I'm. Just going to pull out. One here mean time to recovery, which is the one that directly relates and that Netflix is Need for Speed. They noticed in their survey, which is very comprehensive, I mean they gave it eight to ten thousand respondents. So it's credible survey.

B

They found that the difference between high and low performers, measured in percentiles was nearly fifty times in 2014, the 48x on the right and that's already very big, but it's grown even more I mean in the year between 2014 2015 surveys that went out in 68 X and you know, look at the other numbers as well. It's just extraordinary the difference between people who are on the fast path and people who are sort of stumbling along, and it's the difference between a company that can make several hundred changes to its systems.

B

You know per day across the whole company and a company that has to make you know, has to have months of meetings just to make one change to the color of a homepage.

B

So what is it about collaborative that? We think makes you go fast in this sense, so Eve works. We are a clap, Native company. We decided when we built our own product to experiment by dogfooding everything in being very fab native ourselves, and so I'll tell you a little bit about that next. In order to make this very concrete.

B

Here's our product we've cloud, which is about helping people, build our native apps and there's a nice screenshot of it on the right lots of different components for deployment and monitoring and so forth, and here's a slightly old pitch dreamer in architecture during you, multiple moving parts and as you can see, this is not a traditional web app.

B

There's a several missing several pieces, and this is a system and, to you know, burn together different pieces of functionality and running the cloud and be available and enable us to change it quickly and we've seen turned that into these business requirements. 24/7 is I'm, always supportable available globally, and the team really wanted to focus very hard on application development. They didn't want to spend time being a.m. Isis admins on Amazon or plumbing around with infrastructure. We also felt very strongly as to startup.

B

We had to scale our costs in line with demand, so we wanted to make sure that if a component was demanded and needed to scale, let's say a piece that did monitoring. We could do that without incurring cost in other parts of the system. That might say to visualization of deployment, so we needed a componentize model.

B

We didn't want to spend money, integrating our people, software, so B wants to pick. Software they'll already worked with other things, and we wanted to run our application. We've cloud anywhere right now, it's on Amazon, but we know that we can lift and shift it to other cloud, including data centers behind the firewall. That gives us comfort that we can grow our business reasonably. Predictably, as we get more customers, and so what does that mean in terms of the technology? What did you learn from that?

B

I mean this is sort of fairly predictable if you've been watching the space time, but it is worth understanding what what's going on here. The most important thing is: automation, automation, automation, automation, and this is end to end from the developers. Your fingertips, all the way through to alerts on the operations screen, also mate. All the things- and this means technology is like CICP, continuous integration and most employment, see is in quite widespread use. Already. Cd is becoming more of a thing now orchestration, which I was personally very skeptical about a few years ago.

B

Thinking that many applications would rather do it manually, but as we've become more comfortable using technologies like kubernetes, Docs or messy we've realized. There is tremendous benefit automating how an app is orchestrated and observability, which is all about automating, different components of the logging and monitoring stack in order to simplify how you can control it and get warned of things before they happen if possible.

B

The second thing that you need is, as I mentioned, really the ability to focus on the app, not the infrastructure, because of the importance of not getting distracted by administrative concerns, but also portability, to do that. You need standard packaging that just works consistently anywhere and right now, today, 2017 the leading contender for that technology is containers, of course, docker containers and other sorts of containers.

B

It may change in the future to use new kernels or who knows what else or functions, but the main thing is that containers are key today and finally, we need to understand a lot of new patterns of the type pioneered by necklace of years ago for doing your monitoring, logging up time, management and so on, and this is all about micro services. There are other patterns, but I. Think microservices is the. Is the big name, the big-ticket pattern for these kind of cloud deployments? It's gonna take a pause mark. How we doing we're.

A

Good, we got a couple of questions in session, just such Dever and I'm probably pronouncing that wrong asked. So. Could we then say the cloud native is a de facto standard for cloud paths as in platform, as-a-service.

B

Let me go back to this picture. I think path is often associated with 12 factor models like Heroku, so I think you know. Cloud native is about a more general set of happens patterns for scaling and making your applications more available, regardless of whether they're twelve factor web apps or even it could be data streams processing down here on the bottom left in our picture we have a bunch of monitoring technology going on, which is by no means traditionally associated with pass.

B

But if you, if you free your mind of powers as traditionally understood and I, think it is, it is the basis of the next generation of platforms. So in that sense, I would say yes.

A

There's a second question here and I'm gonna mispronounce, the name again: I apologize needs attending Stricker Erin doula asks. So how, then, is it that we've can run anywhere? They think Erika wants to dive down a bit into the details. He said he using hash. She caught tools for this, or is it just because of containers.

B

It's we are easy ins, a Shi, Corp tools, we're all still using containers, we're also using communities we're also using Prometheus. What are those things got in common? They are open-source tools that we can deploy where, where you want to and have a relatively low cost of integration. That is what gives you a degree of portability, contain really fundamental to this, because you can package any software on it. It consistently pretty much anywhere.

B

I think that's that that's the secret recipe, of course running things running the same code in two different places, there's not guarantee it will work the same way in two different places as people found in the 90s with Java. You could have a big app that ran happily on the server and just would fail to run on a laptop or worst case of phone, but you know modulo those kinds of obvious changes, I think you're. Looking for sort of similar functionality.

A

Excellent. Thank you same. That's the questions for now. I'll, probably interrupt you a little bit later when more coming. Let's.

B

Move on then, so just to drill into those three points a little bit more I. Think of all of the motion in terms of this ABCDE mnemonic apps, a a for apps and built-ins eschatologically and automatically builds using a CI in cases we've works. We like to use circle for our builds, doesn't mean that it's the best one.

B

It's the one where works for us, that container images are generated from this process and land in a container repo such as key, and then there is a deployment step which takes things out, but the image, repo and dumps them into an execution environment of your choice. In our case, we're using we flux for deployment and using communities for our primary execution environment for our weave out, and we don't only do this ourselves whether we think other people should do this too.

B

They might use different components, but they would ultimately be doing the same end-to-end automation to their deployment.

B

What lessons learn and what else I mentioned tools. So we've already had some questions from it from the audience, but open source is key to tools here, I'm, not saying that the tools provided by providers are not good I'm, just saying that they are ones which won't run in every cloud or behind the firewall. So you need to take into account if you using things like sqs on Amazon.

B

You need to be aware that if you want to run somewhere else, you'll need to think of an alternative, some kind of really amazing tools that you can end up with apps that have like attributes like they use BigTable on Google cloud, for example, and benefit from that, and they might use lambda on Amazon and benefit from that.

B

So I'm not saying that it's a must-have that every single aspect of your application should work exactly the same on every cloud you might have different flavors, but I think I, hope you can see the gist is that is to provide it. You want your business to work everywhere. That's that's the key goal.

B

And I mentioned this before, but another lesson is, you know you really want the infrastructure to be boring. That means it might be exciting for the people that run it, but as the app provider you just want the infrastructure to work, and you want to be abstracted from as much as you can so using containers is a good idea. In some cases you could be using an off-the-shelf pass or a container as a service capability like google, kubernetes engine or Amazon, UCS and I would urge people came out this to watch.

B

Apple I call the one percent failure problem, which is when you have a problem which only occurs to 1% of the time and when you're small scale that as a matter, because you know if it happens, one in other times you no big deal. You can usually fix it yourself, but if you've got you know, 10 million users and you have a 1 percent failure problem.

B

That means you have 100,000 unhappy users at any given moment, which is not at all what you want if you're trying to run a business so think about that, when you consider your tool choice here, what it means to be truly boring and then another lesson learnt is you know we need good patterns. We definitely have people in our team who understand these patterns, but we had to learn stuff ourselves and I. Think it's really important to continue to the journey of sharing.

B

What one has learned with people who work to do it so learning of Matt micro-services and will Micra lifts, which are tiny apps that are not little mom, eliciting a very small, the favors cattle more pets pattern where you can shoot any part of your app and it'll carry on going, because it's a it's just cattle and the idea that observability in control or monitoring, if you prefer a rebate into every part of the application from day one, including in the development lifecycle and other patterns as well.

B

And so in summary, I would like to assert that, above all else, cloud native is a set of patterns and just take a pause there for mark to see if you've got any further questions.

A

There are no other questions, some people asking about if they'll get the slides later and stuff like that, but right now no more questions about the talk.

B

Absolutely the slides will be public and they are Google, slides, so you'll be able to come and comment on them and share them a copy them and whatever you like. Actually.

A

One just one just came in, which is: could you shed some light on best practices that are used, we've around the set up of development machines, instinct or similar to production, so asking about deaf, odd parity and how you are you how you manage their well.

B

I'm sitting here in the office right now with our developer experience team, including Aero developer from Twitter Gilead Demetri chenko, he's actually working on that right now and we will be sharing some blogs ourselves in the future. So I'm not going to give you a little verbal brain, domini I will say we're trying to be as open as we can about talking about what we do in the hope that other people find it useful. So I do look at our blog and also partner blogs, we've written quite a lot of stuff about communities and Prometheus.

B

Recently, for example,.

A

Well, that's great. What I'll do is I'll drop the link to the weave blog into the chat now, so you can check that out, stricka-- great and also one more coming in now, Tom Wilkie says we use a terraform module to ensure the infra is consistent. Oh that's, just a tip. That's be used for a cut so follow up with Tom about that you may proceed Alexis. Thank.

B

You thank you so his picture of Adrian, who used to work at Netflix Adrian's a very good purpose and to learn patterns from he, has done millions of fantastic presentations on this topic up and down the country in all around the world over him and looking at his SlideShare page.

B

So go back to the clan native computing foundation. What is it and what does it do? It is open source to file computing from applications non infrastructure and we are curating and promoting a trusted toolkit called modern architect about patterns. That's what cloud native is what about the CNC F limit. If cloud native is patterns, then the CN CF is tools and education.

B

So, let's talk about that now patterns for what well, hopefully patterns, so that everybody who wants the dearness can do it quickly and easily without having to learn from scratch. Things like availability, automation, acceleration and here are some things that I've mentioned already that these things are I believe all related to important patterns and cloud native and need to be documented, describes and learnt about. If you want to do this seriously, the patterns also benefit for having stuffed where you know.

B

Isn't it nice when, instead of having to learn about something, we're reading a book from them to ended and implement it yourself, you can just get a piece of software and use it. We had a constant question earlier about hash. For me, a very fine product called console which helps you do service discovery, which is an important pattern in inferring sample micro services.

B

Now at CN CF, we are trying to identify and promote tools. You can trust as part of our toolkit, and here are the ones that we've already looked at and brought into the foundation or own process of doing so I've combined ATS, which is a container orchestration platform, Premiere Hughes, which does monitoring and some kinds of analysis D, which is a log forwarding tool useful.

B

So if you're doing long, aggregation open tracing, which is a toolkit for doing interoperable tracing a linker D, which is going to be presented in the next CN CF webinar traffic management, routing proxies and we're in the process of voting on geography and core DNS, which are respectively, transport technologies and DNS technology- and there are many more that have come forward and hopefully more to come. Well, though, it looks like there's no overlap here.

B

We do actually probably have perfectly to have multiple tools that do the same thing, because we do not believe that we are qualified or smart enough to identify which tool is going to win so Dave price visits, the occasional duplication in the future.

B

What might the future look like here is a rubber Marvis picture by the CN, CF and red point and amplified couple of theses which we are using to show people what things might look like in the future. Now some things are out of scope, so you know we're not planning to have any of the things in the bottom layer, for example in the CN CF, and all of these projects are projects that may already have a home.

B

So we're not either we're not also suggesting that the CNCs job is complete when all of these badges are somehow associated with us, but we're showing you what the potential landscape looks like for cloud native, but I hope you agree.

B

It is quite a big world out there that the one needs to understand that the bit we're most interested in you know is this: in the middle castrations on time, some positioning and on the right platforms and observability I think we'll also probably see some out app death at depth coming in at the top and see ICD, but that's more of a new idea. I should just let me just go back to this picture. Anybody was looking at this in more detail, contribute to it. I can go to the github page shown at the bottom.

B

Cn CF landscape by a pull request may be used in order to making changes and here's that rihanna's a simplified reference architecture. There's some people like this kind of thing, there's a whole reference architecture. You can go, look at I've got it in an appendix at the end of these slides as well, and you know it breaks out into different key areas, and this is you know how the management layer breaks out. It should respond roughly to the thing in the landscape.

B

You saw earlier said some trying to separate concerns there to show people what we care about. So that's the software in the tools which are there to implement the pattern so that people can go faster. Why on earth do we also need a foundation? That's a good question.

B

I think what is it Mike is sorry to interrupt.

A

Alexis but there's a question which is very person to what you're about to talk about I, think, which is that strikeout and one of the guys asked a question earlier asked, is the plan of the CNC f hosting projects to become something like the Apache foundation in the future? Is that gonna be answered in the following slides or do you wanna jump into that specifically I? Think.

B

It's it's perfectly reasonable to call the CNC, F and Apache of cloud native, so with some slightly different rules, because Apache has grown based on a set of assumptions about what projects needed in the 90s when it was founded in the two thousands when they got going and it's grown to a large size since there and I think CNC F has an opportunity to by focusing on a specific area which is the clown, a leaf area.

B

Then family, a project that had an inter operate, which is some of the things that not all the patches are those that are relevant to a particular purpose rather than disparate, and that are governed by slightly different set of assumptions about how projects work. Maybe it's supported by different set of tools and services, but if you're, if you're, stepping away from that level of sort of detail for developers and people who run projects for customers. Yes, it's like an Apache of cloud native Apache has been very successful with web and big data.

B

We would love to be as successful with laminate it.

A

There is one more question which is also related coming from an anonymous viewer. How are you going to avoid the problems of the OpenStack foundation? Yes, sorry, the OpenStack is encountered as a foundation. How are you going to make sure you don't become a big vendor echo-chamber.

B

Well, that's an excellent question. I think you should ask yourself the same question again as we go through this set last section, but one of the key differences is. We are much more strongly empowering individual projects and we're much more strongly empowering the community and I think I, don't think.

B

Openstack is now doing wrong, but I think they went through a stage of the being a bit too much kind of vendor to developer in the mix and that sort of skew things for a while I think they realize that and if some all the way from that the project is now more healthy, but I do think it.

B

You know really difference technically is open status, trying to be a grand vision, a single project, whereas CN CF is much more than loose loosely affiliated set of things that work together, not all of which must be used at the same time. So I think that will make it a lot easier, I think the things to to work. In that sense, it's more like Apache, which is quite healthy and.

A

There's one more question coming in just before you jump into the next section, which is on a slightly different tack. How let me just?

A

How would you apply cloud native to service world is it? Is it all, encompassing is a different well.

B

That's a good question to you so and if we go back to the landscape there it is, you know the serverless would be another box in the app def and development row at the top, so I think there's going to be multiple special-purpose, app, dev environments of which service is one and now here by server. Let's find, assuming you mean functions as a service like lavender and not just the business model of charging per function. Call because that is coincidental it's serving us today, but is by no means tied to the current model.

B

The event based model now I envisage then actually, on top of this substrate of orchestration tools, but commodity stocker, swarm metals, nomads etcetera. There will be 20 to 30 or 40 different special-purpose, app dev environments, all of which solve different problems really really well, including being equally verticalized. Things like functions to service, but also potentially business verticals like healthcare, for example, or something else IOT so I hope that answers your question.

B

We've already had a couple of teams put their name forward for service in CN, CF and I expect that it's only a matter of time before we see more.

A

Thank you very much. That's all the questions for now so.

B

What is a foundation? Those of you who are old enough to remember Isaac Asimov, might have read his foundation series in which the foundation is a bit Apple and technocracy run by civilized robots. Now, personally, I think this is a pretty good description of many foundations, but it isn't really the primary goal.

B

Is it a Federation of collaboration powers? Then it's also the case. I mean one of the things that makes foundations more important in more numerous today is where, as Apache came about largely to help us to collaborate, I think the foundations of the modern era also take fully into account that big companies are collaborating here as well, both on the provider side, the vendor side, like you, know, IBM Cisco Google in the case of CN CF, and on the end user side.

B

Companies like eBay for example and you'll, see that so you'll see this more in the future, because big companies are now much more willing to contribute to open source and in some cases it's become a differentiator for acting developers for future employees. And so how do you do that? How do you deal with IP? How a Nydia were the cost of development and sort of amortization of accounting philosophies, that's something where foundations can really really help, because they create a clear legal structure and thereby clear financial structure how that works.

B

And, finally, is it a conflicts with karmic forces, and the answer is yes. It is also that, because we care about a bunch of things like Nokia, now I'm going to talk about in a second and developers, productivity inconvenience, so the Linux Foundation is one of the best name, foundations and I think may even be the oldest. Although I think it's just a bit golden the Apache foundation, I believe in its primary job.

B

It's the safeguard Linux for the long term and provide a Nexus Commons, the collaboration and trust around Linux, and also promote it as a concept and brand which I think it's been immensely successful. You probably saw a couple years ago, Microsoft join the Linux Foundation, which would have been unheard of a decade ago, inconceivable, and so it wasn't a surprise that when the CNC F decided to form what it is, what it did was.

B

It became a jury of the Umbrella Linux Foundation, so the CNC F inherits all of the legal apparatus and kind of bedrock proven a foundation and capability on the LF, but can have its own set of. If you like business and development rules, which allows it to focus on what problems we want to solve in Flour, they C, which include you, know, high velocity development.

B

And so the CNC F really arose by bunch of companies. Will core OS? We've works, docker Cisco, saying: let's do this together and we thought that was a very good idea, and so that was for just over a year ago, in the summer of 2015.

B

Ok got a question mark.

A

Yeah so again, I'm gonna mispronounce. This I think it's Rader. Bouncer Benzer asks what are the conditions or criteria to be included in the landscape, that you were just showing a few slides back, I think.

B

You've got to be able to use github to file a pull request. I think that you've got to convince the the committers who are Dan, Khan who's, the executive director of the CNC F and the to venture capitalist named Scott, Rainey and Lenny press respectively at red point and amplify that your project is, you know, sufficient merit to get in there now in the future, they're planning to have slightly more formal rules than that, like you know something that represents.

B

The project has been around for a little while has a reasonable number of users, but at the moment it's still rather at a hawk, so I would encourage you to just sort of throw your hat into the ring. Do your best and don't be too annoyed if you get a bit of pushback, but it is, it is. The aim is to have a fair and open mechanism for just showing people what things look like right now.

A

Justin Garrison's, just posted a link to a Google, was the word survey that apparently people are filling out around their usage around cloud native right now. I would encourage you to go and put your information in there or to read it and see what other people are doing. Another question coming in how is the CNC F related to OCI ah good.

B

Question so before I answer that I just want to say Justin is writing a book on cloud maker.

B

So if anybody wants to help him with that, I know that he's looking for examples, thoughts, contributors, all kinds of things- go, go them up on the cloud native slack, please so ocre OCI is another sub foundation of the Linux Foundation and there's a really really key principle in OCI and CNCs Oh OCR is aiming specifically to create a standard around container technology, a bit like HTTP or TCP before it, and the ITF is a well-known status body for getting Network standards together and internet standards. I think it should be seen in that light.

B

It means it has very clear deliverables around having a finished one, zero specification that everybody can then go away and implement. So it's very very important to get that right. So you know, as it gets closer and closer and closer to being finished, you know the arguments kick and a more and more intricate and complicated, but the end result, if everybody has enough common, will will be to have a spec that everybody can say.

B

It is what a container actually is, at least in mono version, maybe in the future 200 who knows where a CNC F is not about standards CNC. If it's about projects of Education and that's a different set of criteria, so we actually encourage people to just move as fast as they can and work with one another, and we don't encourage people to try to standardize prematurely.

B

We don't inform committees to hammer out disagreements, although there are work groups forming to try and understand patterns, but we're not with the s. Worse. Is this essentially and.

A

The the chat is blowing up right now, but no specific questions, so you can continue and I'll probably interrupt you in a few slides. Well.

B

I mean if you guys want to break fun. Have your anything, that's absolutely fine. I, just.

A

We don't need prettyboy, you've just been the catalyst for a good discussion. I think.

B

We have got to go so maybe making lunch at the end. So you know common open-source. Is the mission here coming out of this idea of? Let's doing it? Do it together and it's a it means something: it's not proprietary. You know big and small organizations coming together to do things together. Almost this important. It's because of that Marc Andreessen quote from that picture, I showed earlier.

B

Software is eating the world and oh, if you haven't, noticed open sources, it's been editing software, but something I'll, spinning, open source and that's cloud- and you know we all know and love the wonderful cloud providers Amazon as you're at Google, but it is a little bit worrying that you know there could be a world in which that they are the only things that exist. So we would love to be able to have a little bit of freedom in that world.

B

So we think open-source is part of that and without accumbens we risk potentially can lock in which could be scary if you're, a small furry animal or a or a company- and so here are some things about foundations. That I think were important. The great open source to ameliorate lock-in I put stop in the headline. Eggs ameliorate was too long and wouldn't fit, but you know they are there to help people ever have a root, be a plan B.

B

Now that everybody's got involved in open-source it's this is this really does make sense that we should share the boring infrastructure, and- and you know, the big cloud providers are going to run this for us, so they win. Two foundations provide education. They can basically provide a shared result, pool letting thought leadership, materials and a mechanism for cross-referencing them and essentially surfacing the good stuff and demoting the bad stuff. A bit like Wikimedia does for general information.

B

They can create an implicit or explicit magic trusts and through operations like the TOC, which I'm a member of, we can identify quality and interoperability tools and examples, and we can share all these through the Commons mechanism of the CNC F and the Linux Foundation I think that's very valuable, as people are rather confused right now, so we're our aim is to get rid of that confusion. I believe there is a question from the pipe.

A

Yeah, that's right, so it was quite a relevant question for this slide, so I'm gonna say Signum I dig in asked about education. Where can we find the best resources for cloud native education.

B

I've mentioned a few already I mean I. Think the slide decks. That I pointed out in Adrienne. Croft is a sort of generator of these. It's good, the project pages, so the CNC that I'll page is a good place to start and the sub projects, cumin, Eddie's, Prometheus, etc, should go and just dig around in all of these things, all the participating companies like we've, worked some core us and are keen to you and UK you as well. So you look at their websites.

B

I think as time goes by the CN CF website will grow and grow and become more kind of encyclopedia for this stuff, but today it is rather disparate I.

A

Think it's worth mentioning in there as well that also. We have, for example, cloud native Khan, which is a fantastic place to come, as we talked about this kind of stuff. Next one's coming up in Berlin surgery in new Europe. That is convenient, otherwise we'll be one in the u.s. at the end of the year and obviously these webinars and there is also a fairly large meetup network around the world.

A

Now, if you were to go to meetup.com and search for cloud mates, if you'd find that I covering the exact number but hundreds so there's probably one near you where you can go meet people and talk about this kind of stuff, mm-hmm, no more questions right now, Alexis, you can continue.

B

Okay, yeah I mean also, we forgot to mention slack channels, I mean that's. Those are very lively places to learn stuff, okay, so um the last thing is family support, open source projects. So you know project should speak for themselves and then we can promote them. I think that's really important, and then we can make sure they play nicely with one another which they're doing really voluntarily, which is great and talk about. Why that helps users as well to build travel apps faster.

B

So those are the three areas. I think foundations can really help bring this all to life, and somebody's already mentioned the standards. So I'll be fairly brief here, but I just want to be clear that term standards been done properly are slow and they generally emerge slowly or they can be brought out early. But that tends to be, you know, quite a complicated process you should enter that has it has its name with great forethought, what we're trying to do in CN CF. We don't think we need standards as such.

B

We need is interoperability and we need glue we'd like to find conventions that arise from the real world and are in use by the community, and if we see those I can identify them, we can promote them as needed and just any weapon grips and try and find common ground for things like sort of storage and networking and and and so on, but we're trying to avoid just sort of putting a stake in the ground and saying this must be how networking is done.

B

This must be how volumes are done, because I think that that is in a can inhibitive limitation and we're still at least the early stage of its industry, and so that's the sort of picture of that plan.

B

If you will, we started out trying to get moving quickly by bringing in some new projects that are about native and high quality and fast moving, and now we're trying to make them go faster and get wider use, bring them into an overall planet, a framework and story, and over time you know, the community will figure out what the standard tools are on, how they work and there's a blog post about this rift many months ago on just linked to here.

B

If you want to read more about this concept, so just to finish up my section I think this is a really nice way of putting what, when standards are useful in this early stage of the industry. um This is Bob wise, who is a member of the CCF board representing Sampson.

B

He said I call upon the CNC yeah to formally foster a common community contain an implementation project backed by the key Cuban, Eddie's meses and plant family communities. We need a transparent community driven implementation to become the default container of limitation for wide number of motor source orchestration systems. So you know here's here's, somebody calling for agreement between different projects and I.

B

Think one of the roles of the foundation, CN, CF and other foundations, perhaps is to foster that and clarify that agreement in order to how customers move more quickly and build assets, but they want to build without worrying about in the infrastructure anymore.

B

So that's the end of my slides, I believe, there's a couple more questions coming through. That's.

A

Right we're going to go three I'm gonna start with Krishna Kumar. He asks. What's your opinion about legacy applications Moving's cloud native. She just leave them where they are or create new apps I mentioned.

B

Mike relates I think that's an important pattern for a lot of applications. This is when you have a small monolithic application that could benefit from something like active, passive replication. Maybe using some you know advancing trickery or some kind of volume, replication trickery this. This is the kind of thing you do when you containerize, like Monsieur applications, people call this lift and shift I think it's an absolutely critical adoption path I. This is why I said earlier: I'm as a little bit skeptical about orchestration.

B

To begin with, I mean there are so many other things: don't need to sort of all be transplanted onto kubernetes and aqus Worman and DCOs today, but it could benefit some rewriting there also things that don't necessarily need to be written, but it could get a bit of portability a bit of flexibility with just a few few. You know neat little tricks, so yes, I'm not against legacy applications at all.

A

Another question coming in here so sessions that endeavor says and it's a bit of a meta question. I guess: could we say that cloud native is the new name or more probably category from born on cloud architecture.

B

This is the same person who asked if it was the I think.

A

B

Is yes, I think that's a reasonable phrasing, I think born in the clairaut or written as if born in the cloud. So people may not actually be writing these things on Amazon Google or as you or another cloud provider, but they might have that in mind. They might be using something like OpenStack and think that if they've written it for OpenStack, it can be ported to Amazon later or run in both places at once. So I would I would agree with the questioners characterization.

B

A

So one more here coming through, then we have a quick look sidetrack question again from session whenever I google cloud native it blows up with Cloud Foundry. If Cloud Foundry is a standard, then what would happen to other past platforms like Google AWS and as your pass.

B

A

Is a fairly wide question.

B

Yeah I. Don't think that, though, whose passes and go away I mean um you know if you're writing a ruby application on Amazon and you don't use Heroku you've got to have a pretty good reason nowadays. I think the same same can be said of no yes, there are you know: people see that you kind of outgrow her okay after a while, if you reach a certain size and scale, but um you know it isn't most people gonna be running off the Cloud Foundry.

B

You know, Cloud Foundry really came about to give people the the the combination of a cloud native style runtime with the degree of portability with that kind of Heroku software lifecycle, where you just got that push and you can bring all the changes with you. It's kind of really handy I think there are many more patterns than that in cloud native that we're discovering there so I think you know the definition will probably expand faster than cloud families. Implementation can keep up and I.

B

Think town family will be an important but there's a case of cloud native, but by no means the only one.

A

B

A

I'm gonna there's one more question before I: do that and putting the link into the chat to everyone now for the next webinar that's coming up and that's for link a d-link D is the newest edition that we've announced from the CN CF as a project related to that I've got a question for you, Alexis, which I think it's very much in your wheelhouse before we had a question about what criteria are there for getting on the roadmap?

A

Not the roadmap, the name of it again, the big diagram, the landscape, the landscape, there's the word we use. My question is: what are the criteria for being accepted through TOC at the CN CF project?.

B

Wow well, there is some stuff about that on the CN CF dot io website, you can see some previous presentations before people are asked to present they're up there they're up there project between formally on the one of the working group TOC meetings. This is a slideshow style presentation with the QA lasting about 15 or 20 minutes.

B

There is then a period of contemplation, followed by a much more formal proposal in the document and a TOC member has to agree to sponsor that in order to move it fall and then the sort of things we're looking for has production users is moving at high velocity is isn't, is visibly of high quality. You know it feels like it as it future is well. Architected is collaborative stuff like that I mean we have written down some of these criteria, but but those are the kind of the main ones so.

A

We have another question coming in from the Yash Thakur and he says, is golang like the native language for cloud native, because from what he can see, most cloud native tools are built with low-lying these days.

B

Well, we just brought Linh kadhi in as an inception project which has got today go London Asia, which is written on which runs on the JVM and there a lot of good cloud native technologies are out there like some of the Netflix OSS projects and some of the spring projects, as well, of course, written in Java other things in Scala.

B

There are new things coming out and rust. You know what are the advantages of containers. Is you can put pretty much any software into the container, regardless of the language and I think you know it would be wrong to assume that application development in cloud native is tied to a particular language. I think go is clearly the peerless choice of community built container infrastructure today, but I don't think there's something that you should see is a hard and fast decision and.

A

One more now in the Q&A section, which is more of a session such Dever at sharing, thoughts, which perhaps you can comment on he says often I see cloud native is glued to containers and micro services, but he thinks of.

B

A

That's the independent of the underlying infrastructure and can even exist.

B

Well, once again, I find myself agreeing with sash in here. I tried to be other explicit about this in the containers of Micra services. Slide I mean you know, containers are here today they are being used as Americanism for packaging software and then running it in a standardized manner. Such this.

B

The results should be at a certain scale consistent, but there is talk of other types of granularity like that might be enjoyed by you know, kernels or even functions, who knows also that the work container itself has preceded the rise of Alexei and docker and associated with Java containers, and if you talk to Cockroft, he will tell you that Netflix OSS you had container model is basically JVM based still and with micro-services as well. I. Think micro services is an aspect arising, incidentally, kind of cloud native.

B

That is certainly a lot easier to do if you've got clan type infrastructure in front of you, but there are other things like you know: data flow processing, which I think can be called cloud native, which are not micro services. So the lot lots and lots of different planet up happens.

A

Ok, thanks Alexis, so that's just let me do a quick double check. Yeah! That's the last question that we've got so all that's really left for me to do is to say thank you very much. You Alexis for the presentation, I think was very enlightening.

B

Where is I might at monadic on Twitter? Do you want to send abusive notes? Oh yeah.

A

And people are now putting thanks Alexis in the chats, so that's from them to you, I'm putting your Twitter handle in there that's at monadic. If you want to find Alexis and follow him, you can keep up with all the blogs and stuff that he's doing at weaving with the CN CF over there. Last thing, I will say is that this video will be it has been recorded and will be uploaded to to YouTube on the channel that I shared earlier and again.

A

I would encourage you to sign up for link D, because it will give you an it will give you an overview of sort of one of our projects is doing, which is another way of looking at some of the of the work here.