Uma Mukkara, Harness | KubeCon + CloudNativeCon Europe 2023

This is texturing TV. Hey everyone. We're back.

We're here. We're in Amsterdam at kubecon in this beautiful hanger. I want to introduce you to uma mukara.

Uma the last time I interviewed Uma, he was with a chaos engineering firm that had just been acquired by harness and I guess it's what I've been a year now or a little bit more than a year since we join. However, Wednesday your last year, maybe that's exactly correct. So we were just acquired by harness.

I came to Valencia and then we talked about, you know, the reasons for me to join harness, you know, being the creator of well, it must be nice to be here again, and now the litmus product the ks engineering product. It's kind of as an official harness. Release is GA.

Yes. So that's a big milestone. And is it still called litmus within harness or is there a name for it now?

I'm litmus chaos is the open source project that continues to be going by cncf. It is an incubating project. And the Enterprise version for litmus chaos is available on the harness platform.

Harness is basically a platform that provides multiple. Product features that are required for devops, right? For example CI CD feature Flags Cloud cost management and Chaos engineering is one of the modules, right?

What we did after coming to harness is we took litmus as underlying base and integrated totally into the hardest platform and added many more Enterprise capabilities. So now the product is called harness chaos engineering and it's kind of an end to end chaos engineering solution for Enterprises that's powered by litmus chaos, and it's that's what when GA a couple of months ago. Excellent.

Good we got. that way one other Studios. Yes, that's harness that I will.

For you get more information on it there Uma. I want to focus in on chaos engineering though, you know, look chaos engineering Birds on the scene, I guess with the Netflix and Chaos monkey and so forth. It's gotta be eight years ago seven years ago maybe more yeah, it is a little bit more like maybe 11 11 years now, right, but that's how chaos engineering was introduced.

It's been a long ride for chaos, engineering itself and moving to kubernetes the big moment to kubernetes as change the way chaos. The use cases for chaos, right earlier. Chaos engine was mainly for direct things in production.

You're a big guy trying to run orbs, you know, smooth way and you need chaos and you have to be an expert but there's now a natural use case that's brought to you to the four by kubernetes itself. Now you're dealing with tons of micro services. And you want to ship this microservices very very fast.

Right? And you need to keep your developers very productive doing the job right? You don't want your developers to Ship something and then come back and debug because you know, you leak the bug into the production, right?

Sure. So that's that's where chaos engineering can come and help you improve your developed for productivity by running the chaos test in the pipelines. And every time you ship something to the right make sure the reliability of the code is protected by running chaos test in the pipeline.

So now we see a lot of usage of chaos engineering is actually happening on the pipelines. Yeah today. But you know.

There's been a maturity of the thing though too, right? I think a lot of organizations looked at chaos engineering as experimental if you and I don't mean any disrespect. No, but it was experimental not Mission critical kind of stuff, right, you know must have I always like to say if you don't have a must-have you don't have Something I think chaos engineering has.

crossed over to a must-have now it is yes, so it is good to have for SRS running large organizations and for the reasons that I talked about just now. It's kind of a must have for. A fully automated devops shop.

Right? Just imagine you are developers running a lot of debug time. In in something that they could have avoided easily by automating right?

So it's it's kind of becoming a must-have developer tool rather than a good to have large ops tools. Of course, the chaos engineering is still keeping its it's it's relevance to all the SRE and odds, but the way I see it is chaos engineering is creating news is the developer tool itself. That's why you see shift left chaos engineering and there's a new term called continuous resilience, right?

So there is a continuous deployments continuous integration very similar to that you now have a new concept called continuous resilience where developers QA and deployers all and obvious people all are considering resilience as a way to deal with it, right? So definitely it's become a kind of must have for devops. I agree with you now.

Full circle back to what we were talking about you had mentioned that you know harness has become more of a platform with different modules and that chaos. Okay. I said engineering module which is based on litmus is just an not just but he's another module within the harness platform.

Yep. What about integration? Right?

I mean that's precisely the reason why I joined hardness because hardness is a platform. That helps you cover all the Enterprise features. For example, you need to have good or bags and ACLS.

You need to have good authentication a good security measures. So imagine you have a great set of chaos experiments reads and you want to you know, give it off to Enterprises. That they love the technology at the same time.

They need to have the great set of governance features. So the hardness platform provides just that right? So chaos experiments backed by litmus chaos at the same time the Enterprise governance features and authenticating features logging monitoring or dashboarding.

Everything is provided by the harness platform itself, right? That's one and also it integrates well with other harness modules, for example CD continuous deployment. You can now push chaos experiments into the pipelines at the click of Hope button.

You can integrate with the SRE monitoring solution SRM service reliability management solution. You can all so see a good integration with harness future Flags. So now what we're doing is make chaos easy to use in all facets of developer life cycle.

Got it. Fantastic, man. finally What about the show?

That's great. I'm in that. I'd love to come to cubecon and meet people and learn new use cases.

Yesterday we talked about, you know using chaos engineering for Telco grade. There's a lot of adoption of kubernetes by telcos. Now.

They're actually feeling the need for audio manage the resilience, right? So now more of the requests that I'm seeing here is yeah. I know we need chaos engineering.

How do I go about it? Do you have any best practice? It's not anymore about you know, do you need chaos engineering?

So I use this Cube cons to validate some of my thoughts listen to various, you know experts practitioners. It's always great and this time, you know this when you is so big, I see a lot of crowd as well coming and you know speaking to us with the boots and you know in the talk, it's been great. Cuban is always great and again, I get to meet you as well here our budget have you here?

Yeah, I'm just wondering, you know, we have people can't see because the camera's face that way on this side of us. We have the booth by the ranch or suits people and they have soccer balls being kicked. It makes this loud noise.

I don't think it comes out on our things which are mics are good. But it's loud. Anyway, Uma I want to thank you for coming on so it's a pleasure.

No sit. We got to say goodbye. Okay again, you go to harness that IO look up the chaos engineering module.

We're also look if you're into the open source, go check out the litmus project. It's fantastic. We're gonna take a break.

We're live in Amsterdam right near the ranch of Sioux soccer shooting target practice. We'll be back in just a minute. Thank you.

Good to have you.