Austin Parker, Honeycomb | KubeCon + CloudNativeCon North America 2023

This is Textron tv. Just checking here. Hey everyone.

Oh, that we're here. Just a little ad-lib. Anyway, we're live, we're live in Chicago, day two of Cube Kahan.

Really happy to be joined by our friend Austin Parker. Same Austin, different hat. Quite Literally, Quite literally.

Yeah. Literally same STAs this year. So for those who watch Techstrong, Austin's a frequent.

Ve well infrequent. Not as often as we'd like guests on tech drunk tv, but we always love to catch up with Austin when we're at an event where Austin is, and happened to be here at Q Conn this year. So we were able to last.

So him on in Austin. First of all, welcome and thanks for coming on. Thank You so much for having me.

It's a, a pleasure. Always an absolute pleasure. And, uh, definitely made some room the calendar for this chat.

I appreciate you. So, Austin, we're gonna talk about two things today. The first thing I want to talk about is the hat, right?

Oh, it's a new hat. New hat logo. New logo.

It goes with a new job. Yeah. A new, a new, a new Austin, new New Outlook on life.

Yeah. Tell us what's the deal? So Is, again, long time viewers might know, I, uh, was at LightStep for quite a few years.

We got acquired by ServiceNow, and it's a great team over there. ServiceNow builds an amazing product that solves, you know, huge problems for huge companies, but, uh, it's, it's time to make a change. Sometimes you gotta, you gotta get back into the startup life.

And, uh, I looked at what was out there and Honeycomb really aligned with, you know, what I think is important about both as a place to work, but also what's important about Observability and Open Telemetry. And we share a lot of the same values around that. So I started a couple months ago and might have seen it on LinkedIn, and I'm super excited to be honeycombs director of Open Source.

Absolutely. So there you have it, honeycombs, director of Open Source op, uh, Austin Parker. Now, just the same way, Austin's no stranger to us, either is Honeycomb and or neither is Honeycomb.

Um, of course, Christina Charity Majors is a frequent guest on our video series. Liz Funk Jones is of course, very well known out of Honeycomb. Um, they're a great company.

Fantastic. Many ways they, they kind of helped invent the observability, uh, market ecosystem, if you will. What, what kind of, you, you've spoken about an alignment, Austin, what, what you think was the alliance?

Like, what, you know, go a little deeper on that. So I think there's, you know, when you're, especially when you're at a startup, there's always trade-offs, right? And even at big companies, there's trade-offs, but there they're different kinds.

But when I looked around and I looked at, you know, one thing I spent a lot of time researching is what are, you know, what's new in observability? And if you look, there's, I mean, if you look around us right now, there's, um, quite a few, you know, observability tools popping into the market. And a lot of those are built on the strength of a pro of open telemetry, enabling it easy, enabling, you know, universal standard instrumentation for things like metrics, logs and traces.

And that is freed up just so much, um, room for people to, you know, experiment and, and invent new things and innovate without having to make these really big investments into, you know, instrumenting a, a node JSS server for the 14th time today. Right. And what I noticed all of these, a lot of these new, new companies have in common is that they all look a lot like Honeycomb.

Yeah. From a technology perspective. They're bringing together metrics, logs and traces.

They're putting it in, um, a wide column column store database. And so you get infinite cardinality and you can do these, these really interesting queries and you can see what's different and what's similar. And it's like that, that's all a lot like Honeycomb and I, and one of the things I've learned over the years is that one of the biggest indicators, you know, one of the biggest indicators that you've got it right is when you have people copying you.

Right? And I figured, hey, if you're gonna, why mess with the knockoff? Just go straight to the source, right?

So I I'm very, yeah. You know, I, I think they've got a, I've always, and I've said this before, you know, I've always been impressed by what they've built, right? They've built this amazing tool for experts.

I think what is different about them and a lot of people is because they have this strong foundation to build on. You know, we can take that and go towards something that is for everyone, because I think Honeycomb gets a little bit of a bad rap as being like this really complicated, hard to use thing. And it, you know, depending on how you're approaching it, it can be or can't be.

But one, one thing that I, I think we can do better at is being, you know, a product literally for everybody, right? No matter where you are on your understanding, observability your journey towards it. You know, I've, I've read the, you know, charities book, and it's a great book.

Not everyone wants to read a, uh, thesis, three or four books before they, they start using a product, right? So, and this isn't just honeycomb, this is everyone in kind of this space. We, we make it really hard for people to just jump in and find answers to problems.

And I think it's interesting, if you think back to, you know, 10, 15 years ago, companies like New Relic, how did they get started? They got started providing pretty straightforward answers to common questions in things like Ruby applications. And we've lost that a little bit with all of this, like observability pipelines and AI and, uh, secret CA all, all the, the, the breadth of things you can do now has made everything very complex.

And now the tools are almost as complex, complex the thing that you're trying to observe with the tools. So to that end, I think we have a really great proof point of how this can work. So last week we introduced Honeycomb for Kubernetes, which is an opinionated, um, basically a opinionated set of features on top of Honeycomb that let you really quickly and easily get started sending open telemetry data, uh, from your Kubernetes cluster, from your application to Honeycomb.

And then we curate that data. We only give you what you need to know, you know, as an application developer. 'cause there's a lot of, you know, Kubernetes is super complex.

Tons to it. You need, you don't need all that data though to understand how is Kubernetes affecting your app and vice versa. So we curate that data stream, we give you exactly what you need.

We put it in places that are easy for you to access, and we guide you to it. We say, Hey, we see you're using Kubernetes, so we're going to just show you these things. And I think that's an example of how these very powerful platforms can be used to provide these really curated experiences that don't require you to necessarily read, you know, read all these books first.

So I have my own theory surprise, um, Hit me with it. So I, I think when you look at the observability space, right, and, and, and especially as it relates to companies that look similar to Honeycomb, there's a couple of reasons. One is sort of, I call, you know, a parallel evolution.

You got a certain fact pattern, a certain problem you're trying to solve. And if you, I mean, there's a lot of ways to skin the cat, but if you truly do have the best problem solve Yeah. Way to solve the problem, other people are gonna come at it and come to that same conclusion.

Sure. So parallel evolution, secondly is just flat out copycats, right? There's always people, and I've been in Tech Austin, you know, longer than you.

Oh, sure. There's someone who says, I, we could do what they do and just change some of the names to protect the innocent and do it cheaper and better, or at least cheaper. Cheaper.

Yeah. No, I mean, imitations is a serious form of flat, right? Absolutely.

So there, there is that as well. But then I think unique to observability is observability is not, in spite of what you might say, is not an easy problem to solve. Absolutely.

Right? And, and I think one of the big, uh, successes of the CNCF was moving beyond Kubernetes to open source projects like Open Telemetry, like Prometheus, that allowed, allowed the whole community Yeah. Access to these tools that, that gave us, that gave us the insights and the ability to act to, to put our, wrap our heads, arms around the data.

Because observability is a big data issue in many ways. You, you're exactly right. I think that's one of the things.

So it's been a, you know, on the topic of Open Telemetry, um, I've been involved in that since the beginning. I was actually, I never, um, well, I just, uh, was elected to the open Telemetry governance committee, which, you know, we're, we make, make sure every, it's working for end users and contributors alike, right? Yep.

So one of the things that we're doing right now is we're really trying to take a look at like, where are things in the project? And I was talking about this earlier this week at Observability Day, where the things that we set out to do five years ago, we have done, we absolutely, we said we were gonna have metrics, traces, logs, we were gonna deprecate open tracing, deprecate open census, and have this stuff done. Now, at the time we said we'd do it in six months, so it took a little bit longer than that.

However, we've achieved our original goals, and now the question becomes where do we go from here? So I think there's, there's a couple of things that that involves, but one of them, to your point about data, is I think Open Telemetry is really proving to be a Black Swan event for the entire monitoring slash observability industry because it's changed. Previously you would have these very, these vertically integrated, you know, instrumentation, storage query, um, visualization, alerting, all these things at the top, and it was all really bound together.

But now with open symmetry, it knocks it all on its side because now the instrumentation, the actual data itself is down here and it's independent of everything above it. Which means that like the storage can now become sort of independent. The query layer can become independent, the visualization layer can become independent.

And I think the future observability tools looks a lot less like a, you know, a platform where you're just getting every single possible thing you could want in the same, you know, same provider and more, almost like a toolbox of best in class tools that different teams are gonna use. And it's all gonna operate on sort of the shared data layer. Um, this, this data light and how companies are gonna differentiate themselves is gonna be two ways.

One, how they help you manage this tsunami of data that open temperature gives you. And building those curated experiences and figuring out what you need to know and what you can kind of throw away, or you can look at later. And two, building really interesting and unique sort of visualization alerting, um, understandability kind of tooling.

Absolutely. Now it's, uh, it's cool. Uh, I don't wanna make it all about Open Telemetry, but look, it's the second largest projects in, in the CNCF and for good reason.

Yeah. And, um, one of the interesting things about Honeycomb though, you, you spoke about it earlier about it being an original, is it was doing it, it was doing, honeycomb was offering solutions in this area before there was an open telemetry. Yeah, I, I mean it's, I see our, you know, honeycomb has always had a very like, Hey, just send us what you got.

Right? You can have data that works in Honeycomb, even if it's not open Telemetry data. If you have your own internal tool you've built that gives you these wide context rich events with the, with some sort of correlation ID on them, then we'll accept that you send us JSO and we'll take it.

What open to telemetry lets you do is it's those, those curated experiences, right? Because if you're just sending us stuff, we don't have semantic information, you know, we don't have the metadata actually know, like, what is your environment like, what is your architecture like? Open to telemetry gives that kind of structure and we can build a lot of cool stuff on top of that structure.

So we can, A really great example is if you're in Honeycomb, there's a little field on the side and all of the different attributes that you have in your data show up there and you can search through them. And so you see things like K eight, kss, you know, food at Bar Baz, whatever. We can just populate that with the description of what it does that comes with it through Open Telemetry.

So if you're an engineer and you're like, I, I see I'm getting an alert on some thing and I don't know what this does, we just tell you right there, right? And I think that's goes to the point we're not open. Preliminary frees us up.

Um, and, and really the rest of the industry from having to kind of build that base layer and let's us focus on actually providing really good experiences for our customers, um, so that they can build better experiences for their customers. I love it. I want to move away from Open Telemetry, move away even, I wanna come back to your new role.

Alright. Director of open source. Yeah.

What's that mean exactly? So What I'm looking to do is, um, a couple things. You know, honeycomb has always been a big part of the Open Telemetry community.

Uh, but one thing people don't realize is that we've, we're very cloud native, you know, honeycomb runs on Kubernetes. Um, we have a very modern, you know, tech and we rely on a lot of the, you know, open source technology that CNCF, um, has, kind of has its home there. So one thing I'm really interested in doing is giving back almost, right?

We, uh, this is it, it's, eh, we announced it, but it's not like, uh, it's not a secret. So I can tell it, okay, It's not getting in trouble, but We we're not gonna get in trouble. But, so one thing we announced very recently is Honeycomb for open source and we're partnering with the Linux Foundation, uh, any Linux Foundation or CNCF project that needs observability for their project, we will, we are offering complimentary paid honeycomb accounts to, right?

And it's things like that where we can go and we can give back to the people that are helping build the software that powers, you know, the really the world at this point. So the other thing is we're really trying to figure out how we can invest and grow the open telemetry community, grow the cloud native observability community, right? Because the way I look at it, the more, you know, the more of these projects and the more of these tools, you know, your databases, whatnot.

If they're emitting open telemetry data, then they are able to be used with any kind. So we want to help inspire that and we want to help, you know, teach people like, Hey, this is how you do this and this is, you know, why it's beneficial. So, um, and then the, you know, some Austin stuff, some Austin, maybe no more animal crossing DevOps conferences, but, uh, we'll see.

All Right, we'll see what happens. We'll see what happens. Hey man, I'm glad you landed here though.

I think it's a great match. Honeycomb in Austin. Parker, no matter where you are.

You do good stuff, man. Thank you. Always a pleasure.

Austin Parker, director of open source at Honeycomb. What board member of of uh, Oh uh, open Telemetry Governance committee. Governance Committee of Open Telemetry Governance.

We're here live in Chicago. We're gonna wrap up day too shortly, so hang in there. We'll be right back.