Open Source PaaS – Catherine McGarvey & Chris Clark, Cloud Foundry Foundation

This is texturing TV. Hey guys. Thanks for the throw.

We're here with Catherine McGarvey who just joined the cloud Foundry Foundation Advisory Board and we have Chris Clark who's a long time member of that program and has been contributing to that Community for every day and we're gonna get caught up on what's going on with the cloud Foundry pass and the age of cloud native Catherine. Welcome the show. Thanks.

Thanks for having me Mike and Chris. Thank you for coming. Thank you, Michael.

All right. So let's start with Catherine. What is the current state of the art for developing Cloud native applications these days because we've been at it for a while, but everybody will tell you that building.

These AppSec is not necessarily easy Cloud foundry's been around for a long time. So are we going to finally get our cake and eat it too? What's going on?

Yeah, great question. So I think what's been really great about the cloud Foundry Community is from its early days. It really understood the importance of getting the developer experience right and invested heavily on you know, I don't care where it runs run it for me and this Concepts it's not applies.

Even if out that infrastructure that foundational layer changes with kubernetes. There's still a lot of interest in making simple foot developers to get to that deployment State and doing it in a way where they don't have to know every layer underneath their application to run it well and that's where I think the cloud Foundry community and it's a best bit with Cory and in continuing to really understand developers and work. We close closely with them really understands this concept and sees the potential for continuing to invest in this space.

Chris if we went back in time, one of the things that people always felt about the cloud Foundry pass is that it was Heavy. There were a lot of components a lot of virtual machines to install basically, you needed a cloud service provided to kind of run it for you. At least for most people that's how they approached it.

Where are we now? Are we going to see more people deploying the past themselves or is it always going to be consumed as a service or just you know, what are the resources required and can mere mortals approach this thing now, right? Yeah.

It's a good question Cloud Foundry has traditionally been a very opinionated heavyweight platform, right and That provided a lot of value right? They did it did. Almost everything you might want it to do for application developers and for operators.

However, it does that made it a little more challenging to be interoperable with new technologies. And you know, there's there's pros and cons to to kind of trying to do everything certainly then kubernetes kind of upended the entire ecosystem and And so there's been a few efforts to make Cloud Foundry and kubernetes play nice together. And the latest one karifi is a much lighter weight implementation of the cloud Foundry CLI and API and it's back entirely by kubernetes custom report resource that's definitions.

It's fully kubernetes native. It's still pretty young. But but it's actually quite easy to deploy yourself.

You can run in a laptop. It's it's not nearly as Yeah monolithic as the traditional Cloud Foundry infrastructure has has been in the past. So having said that you know, there's still a ton of people using the old Cloud Foundry on VMS and it still works great and it's production ready and the current customers are using it even more than they were in the past.

So that's not going anywhere either at the same time. Catherine everybody I talk to will say that their platform is developer friendly and they've been rolling out application development platforms for on kubernetes for several years now, but yeah take is still. Yeah.

I think we measure the number of developers that can write applications in kubernetes in the you know, maybe let's say a million plus maybe two million out of a total pool of I don't know 40 million. I'm just kind of around it up some numbers. What is the real challenge in building a developer friendly platform for kubernetes?

And why haven't we not gotten there after all this time just yet. this question I could sure you hear a lot of people share that God are Developer friendly platform and the real sweet spot with the developer friendly platform is not just the time to value for developers to actually get up and running and understand the ecosystem and sort of limit how much in-depth knowledge they need but also the collaboration across the company with the security reviews the application operators everyone else that's involved in securing applications and to be able to do that in a collaborative way without a ticket-based system you think about developer friendly? It's not just developers that matter in that it's all of the others that go through that review process to get that app in production and having them on board with it.

I think sometimes to be stopped developer friendly. We make it easy to get started, but we don't make it easy to get to production. And Chris Catherine brings up an interesting point because we have seen people build out these devops platforms and they wind up stitching all this stuff together and I guess my question is and this is do Enterprises really want to do that or they just want to write code ultimately and maybe we're trying to get past all that heavy lifting.

Yeah, you know kubernetes is fairly user friendly to get started with. Having said that then yes, you have to if you if you want a production Enterprise ready platform for your developers. You have to build it yourself right now.

if you want it to be kubernetes native, which obviously a lot of companies are now reinvesting are focusing more on that and for I think good reason so there's a lot of value in not having to do that. Right like in Cloud Foundry still has the same value proposition. It's always had that you know, that's done for you.

You don't have to think about that with with traditional Cloud Foundry and now creepy and we're hoping that that becomes, you know a player in the kubernetes, you know platform world. I know lots of people are trying to solve for this right now, so it's gonna be a competitive space but still kubernetes the kind of day two problems that arise once you get set up with it or security and and tracing and logging and metrics and and identity stuff that there's there's a lot that you have to put into building a platform and ultimately you don't want your developers doing that and you probably want to have a lean operations team that can candle all that themselves and not You know, so right now we've got all these companies that are kind of doing the same thing and building their own platform. And that's I think open sources the perfect place for so, you know come up with a solution that can kind of work for everyone.

And so we're trying to do that. Lots of other people are trying to do that, too. Can one of the things you will hear a lot of people focusing on is there's a term Cloud native at least in my opinion.

It seems to be one of the more heavily abused marketing terms these days because everybody's calling their thinking Cloud native, even if it's just running on VMS and I guess my question to use do you think we're approaching some point where people are starting to really understand Cloud native means something that scales up and down easily. It's microservices based and do we really appreciate what the term means these days because it seems like some of our marketing friends of just gone the town and throwing it and everything that moves and that quite everything is really cloud native. That the term really sort of evolved it started off really talking about 12 Factor AppSec and what does it mean to run at scale and run an environments where you can't guarantee 100% uptime.

Do you have failover? Do you have redundancy? Are you managing state in your application?

That's kind of where it started. It's certainly evolved to the point where we're talking about. Does it run on a cloud and I think that the reason those sort of those 12 factors were considered as thinking about these are the best practices to run on a cloud well and to run an environment so that you have resiliency in your application it certainly to your point.

I wonder what the new marketing term will be that sort of evolves us to the next, you know, SOA back in the day. We've sort of seen a few of these waves, but the intent was what are the best practices to make your app run well in these uncertain environments and run well regardless of which platform you're running on that was sort of the intent behind the chair of my belief. You know Chris a lot of people initially Associated Cloud Foundry heavily with pivotal and pivotal became part of VMware and and that led to tanzu and all kinds of interesting things that ultimately came together.

But I wonder if a lot of people are scratching their head and still wondering, you know, his Cloud Foundry two VMware specific or they're enough people contributing to this Beyond VMware. Yeah, fair question and and one that you know that's been around for a bit. Yeah, I mean we're very lucky to have pivotal and now VMware, you know supporting the project as much as it has there's always been an effort to get more parody within that and we did see we had some you know Sousa when they purchased Rancher they went in a different direction.

So we've like that. We haven't moved farther away from that. I wouldn't say but we still you know, we've got a lot of sap is is still incredibly invested in the project.

IBM is still a player and lots of other companies are still very involved in the development and in the community, so we're you know what I like to see other or like larger Enterprise or software companies getting involved in the community more of course, but right now I think we're in a healthy place and you know happy for the contributions from everyone. So well Catherine you want to pick up on that but how does one participate in this community? And you know, do I have to be a rockstar developer or can I help out in other ways?

I love this question. I think one of the things that it's so fantastic about this community is it's open governance and how it operates. It's very transparent on that front.

So to get involved with a lot of people that will jump start and help you. There's a lot of forms where you can ask for help and and sort of understand the code more and it's actually been a really selling point for many of the consumers of cloud Factory. So those that are not contributed using it really valuing being able to see and ask questions on the open source side.

So like in many ways that it may be that the contributors have decreased over time, but we've still think some really healthy discussions about what's being produced. Even if it's not being attributed directly from those outside of the contributor Community as well. But no, you don't have to be a rockstar just get involved the team's super friendly and there's just there's just a lot of healthy interactions going on very very much in a way.

They're trying to say that when anyone's working on it. They're working on it with the cff. That on not with their sort of their company on and that's kind of the way the community is always run.

Chris what do you see here in your experience when organizations first go down this path. What is it? You kind of wish they knew earlier on that.

You see them kind of maybe stumble or trip over or what's your best advice to somebody who's just starting now, if you could just clarify the on what path exactly the well if they're on the path of either building AppSec using the platform or implementing the file. Oh getting into Cloud founder. Yeah.

Well, there's I mean there's really there's two reference deployments. Now. There's the classic Cloud Foundry on VMS and there's there's karefi which is young but but a lot of promise.

The documentation is great for both. A lot of new customers will probably be leading towards a more kubernetes native environment. So we'd expect new Cloud Foundry users to be to jump into griefy having said that the traditional Cloud Foundry user base.

The end users are using using Cloud Foundry even more than they were a few years ago and largely on VMS. To get involved in the community more, you know, there's there's slack Channel and there's mailing lists and it's we recently had a full overhaul of our technical governance. So we moved away from kind of a more bespoke.

pivotal Centric I would say technical governance model that we had had for for some years into a TLC and working group model which you know, we pulled from kubernetes and K native and some other open source communities to kind of put that together and I think that's made this it a lot more welcoming for contributors and for people that want to get involved in the project itself. So yeah, I mean, I think ultimately there's there's a number of ways to get involved in the community. And right now they're currently two ways deploy Cloud Foundry and and for different users that they'll have different benefits.

All right, Catherine. Do you think that we will see an explosion of these types AppSec because we finally reaching some level of critical mass. And you know, our Enterprise is making this turn and ultimately who do you see making the decision as to what platform to employ when in these organizations?

A new app development. We're certainly seeing a lot more of their Cloud native focus and I think the challenge for some of these larger organizations is they still have a lot of Legacy applications or if they've updated a part of the ecosystem. There's that part of the ecosystem still has to speak to some more of the Legacy systems which does make bit of adoption and Broad changes heart, but the real driver I think for having a great developer experience are the leaders of developers.

So these the lines of business in larger companies tend to drive it rather than it and platforms because it's something that you want your users to love that sort of Sparks the interest in in Cloud Foundry in getting in in getting into that CF push experience and then I think what tends to happen if the it teams then go. Oh, okay, and this is not that easy not that hard to run. This is how it works to support and then getting engaged on the operation side, but it's typically still seen as a line of business engagement really people that are managing and wanting to develop Quickly tend to be the group.

That seems most interested. Alright, well folks you heard it here first. There's always going to be technical debt to be paid somewhere along the path.

But the issue is how do we live with it going forward while we continue to build some cool new things. Hey, Chris. Catherine thanks for being on the show.

Thank you. Appreciate it. All right back to you guys in the studio.