Haggai Philip Zagury, Tikal Knowledge | Yalla DevOps 2022

This is Textron TV. Hey everyone. We're back here.

We're at yala devops it the Hilton Tel Aviv. It's been an amazing day 700 people. We've had some great sessions some great speakers.

Let me introduce you to my next guest here. He happens to be a speaker at yalla this year a guy Philip ziguri and he is with Tikal noledge. I almost got it to the last word.

Yeah, and I'm gonna call him Philip just so you understand that but his name is a guy Philip. So sorry Philip talk to us a little bit. What will you speaking out at?

Yala devops? Well, actually it's pretty obvious today but platform engineering which is something that's no building up front center here. Yeah.

I mean again if we look at Jay Frog as one of their products is a the j-frock platform. Yeah, but actually if we look around every Company open source or not really sound kind of internal developer platform this means For us as Engineers as maybe a new way of calling what we're doing like the infra team that we used to have is now platform Engineering in a way. Yes, it is.

So it's basically a lot of tools that we've been building, you know throughout the years. So that was my main so, you know, it's funny. The last person I was spark speaking to was from nice nice good here or no, maybe was right in any event.

This is something we're seeing with the evolution and maturity of devops, right? It's one thing look if you got to be like an Enterprise a big company and you have a small team that's adopting devops 12 people, you know, the typical right the Spotify Amazon model two pizzas, right? But two Pizza team you really don't need to worry about Platform per Seekers they use in a few tools.

No, but I think that if you look at you know GitHub actions and GitHub gitlab so their platforms for for their absolutely featured themes, but what's happened in the in real life is that was fine seven years ago. Now there's a bubble here. There's a bubble there, you know, there's several bubbles these bigger companies.

That you can't have five different platforms with 25 different tools. Oh, that's for sure. I mean the problem is when it becomes acute.

I mean, it's one thing if it's in a science experiment the truthfully it's not the cue the problem is the finding and there's a lot of companies here, you know going for the root cause analysis right? It's real pretty real issue. And that's why we need a platform not because I mean the tools that we don't have are lacking.

It's not a problem lacking. Well, it's also a question. I think of uniformity right you can't So we're just in Us in Tel Aviv here at a small conference with seven relatively small conference with 700 people.

They can't seek us the cameras are only facing this way. But this may be a dozen a dozen and a half then yes, and they're all doing a platform. com.

There was seven or eight hundred. companies In the devops race that would you're doing and everyone wants to be a platform. Yeah.

I mean I go around, you know, these conferences like Yukon and kubecon and they're all talking about the same kind of things but I think what's interesting now is if you look at Backstage, which is a Spotify open source product or platform, which is very similar to what Jenkins did about 15 years ago or 10 years ago and I think Jay frog did with artifactory, you know when their tools so there's a lot of you know emerging open source Technologies which even for you know, these two Pizza teams. They can still you know, benefit from those League. Look I I we've had a few interviews here today around open source, and there's no doubt in my mind after you know going what I see and what I do that open source is the the lingua Franca right the one voice the one language That all of these platforms if you want to call them are talking right?

That's the language that developers want today. It's very hard without it to be non-open Sports. I think that open source is a shade or you know this front we put behind the front door, but behind individual contributors, which is really what's going on, but the individual so we can talk about it, but When you say individual contributes, you're not meaning individuals.

You mean companies? No, it's actually it's a role in in companies today and there are a lot of Agents. I mean, I'm a consultant for the last 15 years.

I worked for a company. I don't work for a product company, but I still come to the companies are consultant contribute to their effort but from a technological point of view or technological aspect and I'm not biased was there business objectives. So you're in a different role, right?

You're almost a neutral Observer who your job is to help your customer your clients by recommending what is the best tool of platform for the job for them? But here's the thing most open source projects today actually don't have a lot of code contributors. They have a lot of users.

They have some people who do beta testing code contributors. It's our measure all amount. You're measured by the number of stars you get on GitHub, right?

That's exactly it. But I buy the people who contribute to code and again if these if you look at these Frameworks, I mean when I'm talking about platform engineering it's because we all have these small kids and tools that we carry around. Yeah.

No one has any IP there? Okay. So a platform should really give that you know back to the community and that's the new open source and well that I do right in a lot of times they're actually signing over the it to like the Linux Foundation or cncf or what have you?

I'm not sure I'm you know again, I think that again the cncf all these organizations all have some kind of agenda, which is fine. I mean someone we've got to make a living somehow but I think if we look about we look at you know, the fact that a library contributed to a company we can continue to it doesn't come it doesn't have companies and again backstage is a good example Jenkins artifactory, but let's look you good. Ladies.

Jenkins is now part of C. Up. I'm not sure again.

If the number I mean this I mean again in America, everything is like large and huge so maybe for Israeli like me looks like there's so many organizations. Where do I go? Or do I need to choose between the CDF or the cncf or the Linux Foundation or can I roll the same?

They're all the same, you know father part of that extendation. Yeah, I know but again it's a but I still want I can't go to conferences every day of the year. You know, I know hey, it's good business for me make Keeps the Lights on here.

Yeah, but all kidding aside. You know you the Linux Foundation is a very it's a very weird. It's a foundation of Foundations didn't they kick out?

Let us know just recently no. No Linda's there. I was I where was I I was in Austin at the open source Summit and they had Linus keynoteing it.

So I'm not sure I think he left the kernel or something. I don't care. What's that?

I don't think he's is involved day to day and Colonel development, but he's still has you know that role where he said I think if we go back to our topic of platforms, so you got me going on the other way. He's no but these company I mean these small projects is why we need a platform. I mean if I want to contribute to a nope, it's those project today.

It's very hard to get into it. If I don't know exactly what let's go back to Jenkins because this is a great example of it. So Jenkins we started by Cake Because that's who are we a friend of mine.

He was Chief scientists. Well, he was his son. Then he helped found or his chief scientists the cloud beast.

And for a long time Jenkins was a cloud bees and I know I know gay myself and yeah, I mean the good news is KK and the rest of the Jenkins community did a good job. Of keeping the distance between Cloud bees and Jenkins. Yes more than we've seen in other cases.

I think kubernetes and Google is a good example a very similar one though. because kubernetes early on was given to cncf so that's the real beauty of these foundations is So I've been in open so long time right so early on in open source, you had like Dr. Richard stallman, and the if you ever read the book The Cathedral and the bazaar no read the book.

I will it's a great book. It talks about the early days of Open Source swear. It's kind of the Wild West it was like a bizarre.

Like I'm very I like the history of so read the history it was you know, it was very idealistic in terms of how How it would work right? But then what happens is we entered the next era of Open Source, which I call the big brother error IBM and basically what that would be is look a project was open source, but it really there was one company that owns it that Managed IT. Yeah and that company did it for their own.

I mean, you know, they gave good lip service to the community, but they did it for their own benefit. So what happened was is if you were Cloud bees with Jenkins, it was very hard to get Jay frog or other companies to contribute real code to Jenkins because it was being done for cloud Beast benefit. Jenkins is better than others, but there are others that were even worse.

I think the plugins the plugin interface was a different thing right because I was all third parties. But then we moved to this foundational model for open source Eclipse Foundation Apache Foundation Linux Foundation. Where you could have a whole bunch of vendors commercial vendors as well as end user organizations.

And and end users themselves all contributing to the same. Project Prometheus is a great example amazing light step kubernetes. I mean kubernetes was started obviously by Google but it's been with cncf Linux foundation for a number of years.

Look at the people contributing code. I might not be very popular if this goes, you know life but kubernetes is jwe all over again. Yeah, it is.

It could be the same. Japan and they have very good marketing. And again, it's I think they built a platform which was very similar to what son did with Java.

Yeah, and then Tomcat came up and weblogic and all those, you know add-ons around. Yeah. So look for my two cents kubernetes is hard.

It's one of the hardest I don't understand. How are you by the way technical I mean technical enough to do that. I need is hard as s***, right?

And I like the first time I saw I was at a Doctor Con in Austin and I came back to the office. I said, you know, everyone's going to this kubernetes thing. I don't get it in November 2019.

I wrote this blog post saying what's wrong with swarm. Why do we need kubernetes? Look if we were in Las Vegas and we were making book on what was gonna be the dominant orchestration to look everyone caught for for containers and what became cloud-native kubernetes probably was a third or fourth or fifth best shot after the warm mesos, even my friends at Rancher.

Oh Rancher I think is a great product was a great product it is, you know, it's fun now nobody I mean, yeah, but they don't have their they they've adopted covering that he's they have stuff they've built around okay threes and stuff like that. I use this for training every year. No, it's great product.

And again my friend sharing with the was he was the CEO now, he's just but for whatever reason The war is over, right kubernetes one. That is look at. Yeah, that's for sure.

sometimes this I was a big OS 2 user. I loved OS too. This is I was in the Army and that's what we used to use the taxes and it was beautiful much better than Windows Ah, that's for sure.

But here we are. I forget the name that I always to war. No the guy from the docker Solomon hikes.

He's darker. Oh Salomon, one of these guys in the development is also chubby guy like us with a hat used to come. I forget his name.

We were in the conference. I think in with the Jay frog in 2019, I together overnight now we're there. Anyway, I used to say who's gonna run.

Do you see a McDonald's guy in McDonald's running cute cattle. Do you see that happening? And I said no wait.

But it's everywhere. I mean it is now it's excited. I think I think McDonald's are running also, I wouldn't doubt it.

Look, it's you they here today. We're talking about putting kubernetes at the edge kubernetes running iot devices with containers on them. So a perfect example, but you know, what did it have been as successful of Google had kept it.

Just as Google not do you think it's they telling us the whole story? I'm not sure got what I'm not sure the kubernetes is the whole story. You know, I did it.

We did a webinar or another webinar Roundtable on this a couple months ago. Is there what post kubernetes what what's this? I don't know.

I do this I go to start a company is for groups of 20 and 30 and I do usually a three or four day course on kubernetes and one of the question is what's gonna be in fight. What am I going to do in five or ten years time from today? I know idea.

Maybe it would be a phone. That's just everything from the phone or I have no idea what the interface is going to look like. I don't know but when you you know, I interviewed Tim Hawkins, I'm actually We have August 10th.

We have Cloud native a virtual event. We're doing and in that one. I'm interviewing Tim Hawkins who is principal engineer Google.

He was on the original Borg. Which became cool. Yeah board.

Yeah on that team about this question about what's the future look like? I didn't do the interview yet. So I don't know what to tell you.

But so let's see but August 10th you can watch it but I mean but that is an example but it's it's certainly a platform a lot of the whole observability platforms have become that's why I think that we haven't we don't get the whole story. We just got a framework which does you know, the basic stuff like deployment schemes. Hi availability.

That's Again part of this model is everyone starts at that even level and then it's up to what companies do with it from there. I agree. I mean, it's very similar.

I mean when I teach kubernetes I say it's like the amplifier of the colonel. Really go. It's like bringing the colonel to the network, which is something that was hard to do in a single machine or a vax or always do whatever because it was just one CPU eventually.

So these four Primitives we were all running for were only these four parameters, which was a limited amount and today kubernetes just gave this, you know the scale. Yeah and again, right You put the time into it. Yeah, we took us about three four decades, right?

So now but yeah good. Well took over that is to become you know. But when you look at the adoption of kubernetes compared to some of the previous kinds of things it's crazy.

It's a crazy adoption curve. It's crazy and again the ease of use and the because there was so much technology so much. I call it the evolution.

Has come so far. But even you be can actually adopt these Technologies and that's the big and that is that's a direct. That's a big sleep forward.

Yeah. Anyway, we're about out of time Phillip for people want to find out more about your company. com?

com. All right, and thank you talk to thank you for being here. Thank you.

Pleasure being good yala. I think that's gonna wrap up our coverage at yalla. It's gonna fight.

Well, it's almost five o'clock. It's been a great y'allah devops. We're happy to be back in person.

We're happy to be back in Israel. We will see you on regular text Chuck TV this week. Hop back home in Boca Raton until then I'm Alan Shimmel.

We're out.