Monitoring as Code – Hannes Lenke, Checkly

This is Text Strong tv. Hi everyone. Welcome back here to Text Strong tv.

I actually have a, uh, a new person to introduce you to to our text Strong TV audience he's never been on before. His name is s le le Leke. If I mispronounce how you correct, I, I got it.

Okay. Yes. Ha.

Hane, welcome. Welcome to Text Strong tv, and thanks for joining us today. Thanks, Alan.

Thanks for having me. And it's a pleasure to be here, Ex. Absolutely.

It's our pleasure to have you. So Hane, you are the, um, co-founder and CEO of a company called C Checkley, and we're going to get into c Checkley in just a moment. Before we do though, I wanted to ask if you wouldn't mind kind of sharing your own personal journey with our audience.

Uh, how did you come to be the co-founder CEO here? Yeah, sure. I'm absolutely too happy to give you a bit of background on myself, right?

So, um, I actually started my career as a, as a software engineer, right? So decades ago, right? And, uh, you know, later founded a Test object and Test object was a, was a startup, uh, which we founded in 2011, and was, uh, software as a service startup for mobile testing, right?

So where we literally hosted thousands of mobile devices and enabled testers and test automation engineers to run their automated test scale, right? Just making sure that their mobile apps are not free. That was, you know, um, the goal of, of my users back then.

Um, in 2016, I sold that business to Source Labs. Source Labs is, uh, a huge cloud testing, uh, company based in San Francisco, and I was on the ground here in Europe with my team, right? So I became the general manager, um, here in EMEA and led the go-to market efforts here in Europe, middle East, Africa, right?

And mm-hmm. I had a fantastic time there, right? So we grew the team quite significantly, grew the revenue very substantially, uh, within the three years where I was there, and this was really, really great, but also what I've seen, right?

So I've seen that DevOps changed the word, right? And I don't need to tell you, right? Yeah.

Um, and, you know, software nowadays is, you know, uh, built, tested and operated differently than it was, uh, back then, right? So we see much more integrated teams, less menu work, uh, and more automation. And I've seen developers thinking about automation all all day long, right?

So, and if I talk about automation, I'm talking about, you know, uh, operations testing, et cetera. And what this requires is a much more integrated tooling, right? So something which is, uh, close to your actual developer workflow to your C I C D workflow, and I truly believe it needs to be code driven, right?

So with that in mind, after, after I left Sauce Labs, I joined my, uh, co co-founders, Tim, Tim and Timo to build check, uh, 2020. So that's three and a half years ago, right? And our goal is to, you know, unite monitoring and testing on one platform made specifically for developers and modern DevOps teams, right?

So teams which are, uh, applying modern DevOps practices, um, to, to what they're doing on a day-to-day basis, right? And this is where I am today, right? Absolutely.

Excellent. So, sauce Labs is no stranger to our audience. We, we featured Sauce Labs here quite a bit, um, and, and they are, I mean, great scale.

Actually, I'm, I'm thinking back, I, I think I remember when they announced the acquisition of your company. com. Um, but you know, at C Checkley, you guys have sort of been a pioneer or one of the leaders in what, what's come to be known as the monitoring is code space, right?

Look, part of DevOps is everything is code, right? But infrastructure is code testing is code monitoring is code. Um, and, and you guys have a twist on that though, combining, as you mentioned, monitoring with testing.

But before we get into that, let's talk a little bit about monitoring as code. When we say monitoring is code, explain to our audience what, what you mean, Haas? Yeah.

Yeah. So happy to do so, right? So monitoring as code, um, actually means that you allow engineers, DevOps teams, um, to manage and implement their monitoring setups as code, right?

So pretty much takes what you, what you just said, right? So it takes principles from infrastructure as code, but also E two E testing, right? Which was by the way, in repositories, you know, since decade now, since Selenium, um, uh, tests are hosted in, uh, in repositories, and we're applying these principles to monitoring, right?

So we're traditionally, you would have your, um, you know, monitoring, uh, configured manually in a UI where, you know, configure what you, what you actually want to monitor or your test should look like, et cetera. Um, we think that should be sitting in your repository, right? Um, so what does it actually mean?

So it means, first of all, yes, your tests or checks, how we call them, are written as code, right? So that's, um, plain JavaScript type script, um, code, which is, you know, uh, checked in into your repository, but not, not, um, only that, right? So in the end, it also touches all the other configurations, right?

So if you have a monitoring system, you want to get alerted, uh, if things go wrong, right? So you have to configure that somewhere. So you, you might wanna get an email, you want to have a phone call, you want to, you know, in, um, integrate other, um, alerting solutions.

You wanna have dashboards, um, you might wanna run checks, not not on the public cloud, but in your own private networks or private locations, et cetera. And all this should sit in, in your repository, right? So this is what we, uh, try to achieve on a day-to day basis, right?

Mm-hmm. And if you, if you think about that, right? So if it's code, then you could even generate that code, um, on the fly, right?

So let's say you have new a p i endpoint introduced, and, um, uh, you could recognize, okay, there's a new API endpoint introduced, and I want to spin up new checks, uh, for this API endpoint. Or, um, you have a new platform for a specific customer deployed, you want to monitor the automatically, right? So mm-hmm.

It's pretty much setting up, um, and automating your monitoring process. And this is especially beneficial in, you know, in environments where the application changes, um, constantly, right? Or where the infrastructure is changing constantly.

Sure. Right? Um, which Is very common today, right?

I Mean, yes, yes. So everyone wants to move fast, you know, every, everyone is, you know, shipping new features to customers. So, you know, um, that happens, uh, very often, right?

So, and I want to give you a practical example here. So, uh, we are working with a global e-commerce company, and they have more than a hundred online shops worldwide, right? And they wanna monitor them worldwide.

And they were looking for a solution which not only enables, um, their operations team to create and maintain checks, but actually their whole org engineering organization, right? Um, so what they did is they took the monitors, which were configured, um, on, on legacy providers and vendors, and exported them into code. And they're now hosting these monitors next to the actual application code of an online shop, which allows their engineers to, you know, make changes right away if the online shop changes, right?

So, um, monitoring is only as good as you monitors, uh, right? So, and, um, let's imagine an API endpoint is changing, then you want to change your monitor as well. So the developers can do that right away without, you know, ping someone in operations opening a ticket, waiting for weeks, sure.

Et cetera, et cetera. So it speeds up the process quite a lot. And of course, if your monitoring sits in your repository, comes with a lot of additional benefits, right?

Between which we know it, um, from applying DevOps principles to, to other, uh, workflow as well. Got it. Let Johannes, let, let us, let me play devil's advocate a little bit though.

Yes, sir. com here for 10, almost 10 years, going on 10 years. I, well, March will be 10 years, nine years.

Um, one of the things that I think we've seen come out of DevOps sort of an anti-pattern is in, in our move to shift left, we've thrown more and more on the developer's plate. And, and it's kind of, you know, it's not logical in some ways because you, in many organizations, the developer is probably the, the biggest single point of failure. It's your most expensive person on the team, right?

They, they have a lot on there. And, you know, over the years we've moved more testing to them, obviously, right? Developers running automated testing or kicking off automated testing, including security and security testing.

Now, we're moving monitoring combined with testing on the developer too. Is, is, is it, is that destroy, that breaks the camel's back perhaps? Or is it just part of what they're doing anyway?

It's not a, it's not a big lift for them. So, um, I would argue, um, that, you know, and in what part, it's, it's part of what they're doing on a daily basis as well, right? Okay.

So you mentioned testing as is now a responsibility of the, of the actual developer, but I would, you know, if I think about DevOps, I think about a joint combined team, right? So it's not a team only with developers, but it's a team of, you know, people who understand operations, quality assurance, um, auto, uh, engineers, et cetera, et cetera. And they all need a, you know, need to have a, a single pane of glass on, on first of all the application, but also, uh, need to have a common ground on, on what they're actually building and monitoring.

So, you know, you can enable developers to make these changes, as I gave you in that, in that one example from one of our customers. Yeah. But the truth is, um, also ses, um, and part of operations part of the thing too, you know, um, they're part of the team and they understand code nowadays as well, right?

So, um, so, you know, they want to have something which would the whole team understands, right? So that's, that's one part. The other part is, um, since developers are doing testing and writing E two E tests, why would you reuse these E two E tests or end-to-end tests, um, for monitoring as well, right?

Or a subset of that and deploy that, um, to check and then run your monitoring, um, driven by these E two E tests, um, 24 7 Sure. Globally, right? So I believe, um, there, there, there are a lot of benefits, you know, um, not not doing things, uh, manually in the ui, but, but really have, have them version controlled, right?

Um, and these benefits are also, as we know from infrastructure is code, um, applying to, to monitoring this code, right? So version controlling, um, you know, comes with a higher transparency. You can see what you, what you have actually, um, deployed.

You can track changes and maybe even reward to previous versions of, you know, what you have. And then, um, making sure you know that, that your app runs, but, but also that your monitors, um, run and do not alert you because, um, the monitor was wrong, right? Um, if I talk about transparency, um, having things burden controlled means that you see who made what change and when, right?

Which is great, especially larger organizations. And, and, uh, if you want to enable multiple teams, um, to work with an existing setup, right? Um, so that's another major benefit here.

So in the end, what we try to achieve is we're making monitoring a team sport, right? Um, where of course your developers can, again, take more on their shoulders and do monitoring, uh, but also your ss you know, talk, talk the same language, uh, with, with the, with the developers and, and are seeing what's, what's actually happening here. Got it.

Excellent. And, and so I just wanna emphasize this to the audience. So when we talk about monitoring as code helping developers, we're talking about monitoring his code, helping the entire DevOps team.

Yes. Right? And, and all the facets of that.

Haas, I I, we only have a few minutes left here. I want to talk a little bit about C Checkley, how, how, give us the on-ramp, give us the, the glide path for our audience to engage the company and, and look at the solutions. How, how do they get, how do they do that?

Yeah, I'm, I'm very happy, happy to do so. So let me sum up, check you once more, right? So what we're building is a monitoring platform made for developers, but also modern DevOps teams, right?

And, uh, you heard that my background is in testing, right? So I was in, in the industry a decade ago. My co-founders actually are coming from operations, right?

So they, uh, work for, uh, large scale companies doing operations there. And what we're doing with our platform is pretty much taking, you know, automated test scripts, uh, which could have been your, or might be your E two e tests, and deploy them as monitoring and vice versa, right? So you could, um, use, uh, our monitors to, to test, and we enable that on our platform, right?

And this is what we call monitoring as code. And actually we have a, we have a Terraform provider, and Terraform is, is the number one solution for infrastructure as code. But also we have a check c l i command line interface, which enables our customers to get started with without, without an additional, um, tool, right?

And, um, check, um, how we think about the whole process. We, we, we think about it as, you know, code, test, and deploy, right? So in the end, you first code your tests, then use them as tests to ensure that your application is running, um, maybe on staging.

And then, um, if that application is running and is about to be deployed to production, then uh, you would deploy your monitors, which are then constantly hitting, uh, your production systems to make sure everything runs like a Sean, right? And I believe you might have one question here, right? So, you know, you know, uh, wanna already go into it, right?

So what's, what's the difference between APM and, and what we do, right? And I recently talked to c o of a large e-commerce company, and what they do right now is they have a perfect inside out monitoring configured, right? So they monitor, um, their just spaces, um, their HTP errors, their database errors, et cetera.

And the underlying assumption was always, okay, if all these things are green, then the customer experience is perfect, right? And now they realize, okay, this is no longer sufficient, right? They need to understand the customer experience right away.

So, um, and what they're, what they're seeing is, you know, a lot of noise coming from there, um, as a monitoring, uh, solutions where yes, of course this on that one server is, you know, running o uh, out of space. But actually, does it mean anything to the, to the user experience? Should we wake up in the middle of the night because that's the case?

Maybe not, right? If it doesn't, uh, affect the user experience. So, um, what they're looking for is to, to build an outside in monitoring, right?

So, um, to constantly, um, monitor the customer experience from, from the outside and, uh, build dashboards for the single squads, which are then, um, using these dashboards, um, constantly looking at them and get alerted in the middle of the night if the customer experience, you know, drops below a certain bar. So, um, and this perfectly describes what checklist is doing, right? So we're, we're looking at your infrastructure from the outside, um, to understand is your, you know, performance up to a specific bar, um, is your, uh, online shop still working, checkout flow, still working, et cetera.

Um, and I very much like that, you know, example from from this city o. Excellent. I like it too.

We're out of time. Johannes, I've gotta wrap this one up. Congratulations on all you're doing with Cze Lee.

Love to hear more about monitoring as code. We'll watch this as it becomes, you know, a bigger, uh, a bigger item in the, in the market, right? It's your gar as you mentioned.

Gartners already mentioned it. And, uh, we'll, I'm sure we'll be hearing more. Keep up the great work, come back on text Strong TV and, and keep us posted.

Okay? Thank you Ellen. It was great to be here.

My pleasure. Great to have you on. Check out, check Lee.

Uh, we're gonna take a break here on Text Drunk TV now though, and we'll be back in a moment.