Eric Schabell, Chronosphere | KubeCon + CloudNativeCon Europe 2023
Learn about how companies can manage observability to increase efficiency vs. living with incident burnout and escalating costs.
Transcript
This is texturing TV. Hey everyone. We're back here.
It's Friday. It's Amsterdam. Kubecon.
The floor is open and man what a great show. This has been great traffic great engagement with people we're gonna hear more about it. I want to introduce you to Eric chabel.
Eric is with kronosphere. When I I so the first I still remember the first time I found out about chromosphere chronosphere. I think it was one of the co-founders.
Is it Martin who Martin Martin Mao Martin now? Yes, sir. See what right right dude, and he told me what day we're doing and I was like I wanted to reach in my pocket and write a check plus I was like, this is gonna I know this is gonna be good at the time.
It was Martin and a few other young people. Yes, and Yeah, you've come a long way since then, but Eric, welcome. Welcome to techstruck TV.
Thank you so much. I still here. What do you do with kronosphere?
So I'm the technical director of or the directive technical marketing and our developer evangelism team, and I'm sorry I didn't catch the last part the developer evangelist team. Oh great, Excellent Man. Give us an idea.
What's kind of your background before coming to kronosphere? Good question. So I live in Europe, but I'm originally from the states.
Okay. I met a Dutch lady and that brought me over here. I picked up a computer degree got into development.
A lot of Open Source. That was just kind of my draw sure that led to a call from Red Hat and I joined them for 13 and a half years, but I went through the complete explosion in Europe of the growth and what happened until they were Acquired and one of my good friends at X red Hatters introduced me to one of the vice presidents here. Chronosphere said you got to meet this lady.
So I was chatting with her and they were looking for exactly this role to fill and when they started explaining the technology as what happened with you with March and mount the same thing happened to me. I was like, wow, you know something I've been talking about for the last year and a half or so by Cloud dated the flood of David the problems people are having that I'm seeing when I talk to customers and they look like they understood it and with the background of coming out of uber and having done this at scale. I was pretty pretty impressed.
Absolutely. You know what we kind of teased it. But not everyone out here knows kronosphere and they're not all 100% sure about what kronosphere does.
Let's talk. If you don't mind our problem that's right here for absolutely tell them about kronosphere. So chronosphere is a cloud native observability platform.
We're a SAS offering when you say observability. Everybody's observability at some level right but we're talking about at the massive scale that you're seeing people do stuff in the cloud kubernetes containers and the explosion that can happen when your business is very successful. That's all Grand and fine but there's an issue around metrics data Telemetry data all the data that you're collecting you have second generation platforms that are storing this and charging you for that and what you would like to be able to do is in the middle to be able to understand what's coming in and decide Terry pick what I need and pay only for that right?
That's that's the problem the scale goes so quickly because you describe a cloud platform. How you to behave and you don't get any more control after that. So think in the pandemic places like in the US where the the food delivery services started going crazy doordash one of our customers you're doing a production data to here it explodes and all your Telemetry data goes like this.
That's not what I want to be paying for if I'm not using it and and nobody will complain about this if you're getting better customer experiences, if you're you're making money, if you'll be able to read me remediate stuff faster, but because of this explosion what you're trying to do with your dashboards get slower and slower because it's got more and more data to plow through and so we've we've developed a control plane there in the center that allows you to get some knobs and dials and take a look at what's value with you and only the store what's a valuable to you? You know, so this is been like you Eric. I've been in technology a long time.
Right and I still remember. Yeah, I remember the first time I met Martin now, I still remember the first time I saw Splunk. Yeah, it's got to be.
I'm gonna guess 18 years ago. So something like that right that really cool t-shirts that said cool things on them, but forget the t-shirts when I saw that this thing was like glue you could put it anywhere in a collects everything. I was like, I want to collect everything.
I want to know because I'm one of those guys who believes we measure everything and everything could be measured right? Yes, and I so I just want to suck that data in if there's a log I wanted if that I got it because it makes sense right when when they're approaching and they're saying hey, you might need it one day well because we're we're all day to Junkies. That's the that's the nature of this piece.
And then you realize just because I can doesn't mean I should especially when I got to pay for it. So you wind up like a hoarder. Yes.
Basically you get walking the house because there's boxes of data everywhere and I'm paying rent on all of this data. And when we were looking at second generation observability, it wasn't called that it was APM, right? Absolutely.
So we had a controlled environment that we knew what was going to be happening in it. And now when it's Dynamic and somebody explodes your business, which is a good thing don't get it and nobody has a cap on production. They have a cap on you know, Cloud usage as a developer not as production and so it happens and then they're all like, whoa, this is how to control and and so You know this whole Market gave rise to hey, what data should we keep and what data you know and not all data is premium data that goes here maybe there's less premium data that goes there but and then what do we need?
What do we really need? And especially I would imagine and I'm interested in what you're seeing here on the show floor. A lot of the a lot of people on this show floor Eric, they haven't been through cycles that you and I have been through in the economy.
Right? Right. And so we're you know, we're on this cycle where we see layoffs things tightening up a little bit Yeah, and all of a sudden people are saying how can I save a few dollars?
Why don't need all this data. I'm not using a lot of it. Yeah.
How do I how do I tweak what I need? How do I know what I don't need that's gonna be a huge driver for Kronos here. Absolutely.
So one of the things that we try to encourage and and it fits really well in a cncf program like this is that you should you should definitely start out with open source Open Standards, right and preaching that for years also in my previous roles sure what architect doesn't want pluggable components, right? So if you stick with Open Standards of ingestion in our field, so with metrics coming in and Open Standards, that means as your infrastructure skills and you're doing it yourself and you're adding more and more components the downside is is you get to a Tipping Point where there's too many devop resources that are now having to come over here and do Infrastructure and you're not doing your core business fair enough to come to Cronus for you unplug from the Open Standards plug them in here. Nothing's changed get rid of all these resources get back to doing what you're doing and you have a managed platform that you recognize the same protocols the same query language the same dashboards, you're still looking at a browser again to your dashboards.
Nothing should really be that big of a shock. What we do introduce then is the ability to see live ingestion to see live data coming in and decide what do I want to pay to store and it's not gone. It's just being aggregated the way at this point.
So only the valuable things for your business are being used at that time. And if you want to come back and say, oh we're gonna change and add a new service. We do need this piece of data, you can turn that back on or aggregated differently so that it does show up and what you're seeing across.
Our customer base is a 50% savings on data storage 50% on average. Wow, we have examples of customers that are that are up. upper 90% you know, there's just there weren't using that much and they didn't realize it.
No it that's exactly it. You know, I look this is the same sort of dynamic that drives sort of this whole finops. Movement that was seeing right where I mean.
For most people Cloud elasticity is a one-way Road. Up. I was shocked and stunned last year when I was I'll be honest.
I you listen to some of these IDC broadcasts when you're working at home and home office and it was on in the background. I was listening and she threw out a number and she said that in Europe it was about Europe Cloud vendors and Cloud users and they're saying that I don't remember the exact but it was in the 80% tiles but are going to have a finops role by next year by the end of 2023. I almost whip last like what you know, you don't hear them say numbers that High.
I mean, I know that's critical match right? He's responsible for pushing these buttons in the marketplace and putting stuff into my architecture that's generating more metrics more data and more cost eventually. It's the CIO or the CTO, but you know, but I just think it's the nature of the Beast I mean Like I said earlier, we're all data hoarders.
Yeah. Yeah, and and we never delete right? Well the developer doesn't really also I was an AB Dev guy for a long time.
And and one of the last demos I put together at the company. I was playing around with some technology doing some get off sinking to the cloud and stuff and the the end of the week. I think five days.
I got a bill for 800 euros, and I'm like What I was wishing wasn't that much but I didn't think about like I was using a platform for containers and when I pushed that button and spun it up, there's logs and some stuff that they have turned on that. I'm not looking at I don't need but it's on by default. Nobody that is and that is the issue.
But you know what? I don't want to I don't want to blame developers solely. Right, I think there there is this.
Data hoarding if we can measure it measure it kind of mentality that permeates from CIO CTO ciso's on down. Right right the developers of one. Clock Cog in this whole machine, but organizationally we seek to you know, the the default is oh, yeah.
Yeah, turn it on. Turn it on we can we can store that. Yeah, turn it on me and you know, but one of the nice things about kronosphere though is just because it's not front and center and you're paying for it doesn't mean it's gone forever as you mentioned.
Right? Right because you like security you want to do a forensic order after a preach or something. You want to have that data, you know, and you got to keep it for a while because oftentimes we don't discover that was a breach until many months after The breach occurred I'll talk a little bit out at some of my talks about cloud data and some of the pitfalls to watch out for and we were looking at some of the research and and the average retention rate is generally a standard policy for a company and that's the first gut reaction.
Let's keep it for the minimum required to keep it for and they don't specify certain areas of the business. They're like and the average across all these companies is 13 months. But what about my development environment where I did a training for two weeks?
13 miles lose that right, you know why who cares exactly flush it after two and again, that's that's the beauty of kronosphere but that's what you get 80% Saving right and right kinds of things. Um, Eric, let's talk a little bit. It's been a crazy week here 10,000 people.
Yeah, I'm assuming you're at the cronosphere booth talking Everyday People. What are you hearing from them, man? Two things that have really jumped out of me one were sold out here, right?
That's yeah, that's not it. That's probably the third thing person waiting list pretty crazy. Right and that shows the enthusiasm of want to getting back in the floor.
Why don't you back into stuff wanted to talk about open source, and I also think the expansion of the cloud has been pretty pretty prevalent right people to pushing hard. The second thing would be or the that would be three things. So the second thing would be I'm hearing an awful lot about cost.
Right, it's definitely oh, I I think yeah people are looking saying hey savings exploring other stuff. Yeah, and the third thing is when they're showing up the booth for the first time. I mean, I've been doing this for years, right and usually when someone shows up you're trying to convince him to listen to a little bit about what my product is or what I'm trying to do or what we do and sometimes it just coming for this swag or whatever and it is blowing through but I have consistently had people say what do you do tell me about it that kind of thing.
They're interested feels like the minds are open people. Are they are every boot not just no I'm hearing you as well. I think part of it Eric is you know, I've been going to keep cons a long time.
kubecon used to be primarily a not primarily. It was really focused on development developing applications. Yeah, right and and so I'm using Cove and containers.
Yeah, and I you know, I got multi multi-threaded kind of applications or I'm moving from a amount Amano kind of architecture to multi and You know, it was really aimed. How do I make developers lives better, which is a worthy topic. Right?
But what we're hearing more now is the Ops end of things whether you're an SRE or a platform engineer or an architect or you know Ops I'm running right? How do I architect? How do I get the plumbing?
Right? Right and I think it's more and more of an emphasis on that people want to hear about it. Now, we're getting a lot of platform Engineers srees observability managers.
I mean, that's kind of Internet on the product that we have I would assume but you're seeing him Bounce from ours to next to ours where it is a developer stand and it's it's there's an interest all across the board, right and and everybody's looking to figure out how to get a little bit right? We got it tighten it up, but we still got to do what we do. Yes, Eric.
We're about to time. I'd like you to look into that camera tell people if they want to find out more about chronosphere on the web. io.
See you there just like it's spelled on his on his shirt there Eric. Thanks very much. Thank you sir, and enjoyed the rest of kubecon say hello to Martin for me and the rest of the team and good luck.
Thank you, sir. All right, we're live. We're here at Amsterdam Keep Calm.
We're gonna be back in a minute stay tuned.





