Ramon Nissen, RedHat | KubeCon + CloudNativeCon Europe 2023

This is texturing TV. Hello and welcome back to kubecon plus Cloud nativecon Europe. We're here now with Ramon Ramon Neeson, who's from Red Hat and we're talking about this open source conveyor project that red Hat's been driving.

There's been some updates this week Ramon. Welcome to show. Hey, thanks for having me happy to be here.

So walk us through the new updates. What did you guys address? And how popular is this project these days?

Okay. So we made a small change. Well, not not small.

I think it it's important we change what conveyor is. So before it was some sort of umbrella Community with different projects at addressing different problems. We thought after we run these state of the app modernization survey with many organizations out there.

We found out that they were mostly focus on the bringing traditional workloads into kubernetes problem. So we thought we should talk use the community into those scenarios. And basically what we did was make and mayor become an actual thing an actual project.

That can be installed. So this week we have released the conveyor operator. There are not many operators anymore.

It's just one with everything fully integrated fully focused on this application modernization migration scenario. So are you seeing people taking Legacy monolithic applications that ran on a VM and trying to move it over to kubernetes? And because when I talk to people is kind of a debate, do you do that and then break it up into microservices later, or do you break it up into microservices first and then move it?

Yeah, what we're trying to do with our project is a low larger organizations to make those decisions. So basically allow them to manage Go bear in their application landscape surface insights about that landscape and then making form decisions about what should be done so I couldn't say it's an antipattern per se to move a monolithic application into kubernetes. There's plenty of stuff that kubernetes offers out of the box that is difficult to obtain on a traditional Legacy platform.

So that is a definitely a migration pattern that we are seeing but what we're trying to do with our project is to help organizations make those decisions and surface the information necessary to make those decisions. What is the relationship between conveyor and cubert people are using Cube vert to run VMS on top of kubernetes? So are these complimentary things or are these alternative things?

Actually, we have a pretty close relationship. So back in the day when Khmer was just community and we had many projects. We had a project called forklift which was used to automatically migrate VMS tradition from traditional environments to keep bird.

Basically, we have contributed these forklift projects to cupert. So now it's part of kubert itself. So instead of us managing that project in a separate way, we have given it to them and we are likely to have further integration in the future.

Is this an open source project? That's run by you guys or did you contribute it to a Consortium or you're thinking about that or and if it's still run by you guys why and what's the thought process? Yeah.

No, we we kickstarted the thing this thing between a redhead and IBM research and I was in September. I think we got accepted by the cncf as a Sandbox project. So now it's owned by the cncf and we are still the lead contributors to the to the project basically is now owned by the cncf.

So what we are trying to do is bring as many vendors gsis or even organizations on board to make sure that we address all the problems that people are finding are trying to solve out there when it comes to application modernization on migration. How big a community is this right now? If you want people to join it, how would they join?

Where do they get started? io. We also have in the root of the Camaro organization in GitHub.

We have all the coordinates for you to join our community meetings. We have a slight channel in the conveyor. Sorry, the kubernetes is locked.

So there's plenty of ways to reach to us in and join us in terms of size two core contributors right now red hot IBM research. We already have some contributions in the shape of Rules from some Microsoft faults to help you migrate towards AKs and Asher app service. So we're we're just growing I think we have I don't remember the second memory.

It was like said 78 contributors for many organizations. So far core contributors are again right ahead and IBM research. We'd love to welcome anyone.

It seems to me there's a lot more urgency surrounding application modernization these days. What do you think's driving all that? Definitely.

I mean, I think the problem or I wouldn't call a problem with the situation that we're seeing with lots of organizations is that they're adopting kubernetes they are leveraging it but they're leveraging in a very small percentage of their application portfolio. So maybe they have adopted kubernetes. They have maybe 10 to 20% of their applications deployed in there and then they have the remaining 80% of their applications running on traditional infrastructure.

They are kinda learning about the good things that kubernetes and the flexibility that kubernetes offers and they're wondering how to leverage those capabilities with their traditional applications. So I guess that's kind of where these these need is is coming from and also trying to have a unified operating model for all your workloads. So what comes next for conveyor what's you know on your list that to Do's what's at the top?

Okay, so short-term. What we want to do is focus on multi-language support for the analysis module. We have an analysis module that gives you insights about what needs to be changed in your applications right.

Now. It only supports Java. We are building a new analyzer from scratch.

It's an implementation that it's based on the out. Yeah language server protocol technology that made Microsoft is using in Visual Studio code. So we want to abstract the complexity of all languages from the analyzer and make it language agnostic with that.

We want to tackle things like support for DOT net which is a huge thing. net right now to enable these donut workloads to Ronnie and containers within kubernetes. So that's the next day next big thing that we have and on.

Midterm we want to enhance our assessment module to make it a little bit more flexible so far. We're tackling the continuation use case, but we like to talk other cloud-native migration scenarios. And also we want to tackle the problem of assets generation.

So automate the generation of everything. You need to run your application on kubernetes like deployment manifest like em Docker 5, like maybe preliminary pipeline resources all that stuff to make the whole thing as a streamlined and simple as possible. All right.

Any thoughts or maybe adding Cobalt to that list? Yeah, I got asked about that all the time cool things that there is an language server implementation for Cobble. So we should be able to cover if it's an appetite for that in the market.

Right? What's your best advice to folks when they go to modernize an application? What do you see folks doing that just makes you shake your head a little bit and go folks.

Maybe there's a better way of thinking about this. Yeah, the way we think of application modernization on migration. We'd like to take or at least enable organizations to take a holistic approach other application portfolio.

If you just start modernizing application by application without having your portfolio inside, it's very likely that you will make decisions that you might regret in the future. So it's better to identify risks and problems as early as possible come up with suitable litigation strategies and make sure you don't find any surprises later on so my advice will be Try to take an approach that is as holistic as possible. All right, folks.

I think he just said you should look before you leave. Yeah, Ramon. Thanks for being on the show.

Thank you so much. All right, and we'll be back in a minute guys.