GrafanaCON – Ryan McKinley, Grafana Labs

This is Techstrong tv. Hey everyone. Welcome back here to techstrong tv.

You know, the day we re recorded this interview right here, there was a Grafana, uh, conference going on, and we were lucky to snag Ryan McKinley, VP of applications from Grafana Labs, literally right in the morning before the conference really got going to come in and talk to us today about, you know, what's kind of the big news at the conference. And, and one of that is, uh, version 10 of Grafana is, is actually shipping. So, hey, let me introduce you to Ryan McKinley.

Hey Ryan, thank you for joining us. Thanks for having me. I'll actually jump in here.

I'm no, I, I got, I got my title changed. Uh, I Oh, okay. Engineer, which is pretty great cuz I no longer manage anyone.

So, uh, Hey, Is that something Get to work on, on Just Aspire too. So what, what, What's the new title? Distinguished Engineer.

I love those. My kids think I need a monocle, so, uh, uh, Hey, look, you can, in case you know, we need to do a Hogan Heroes kind of thing with Colonel Klan or whatever, it's all good. But, uh, no, you seriously though, distinguished engineer, usually they don't give it out to people who aren't distinguished.

Right. They, it's a, it's an accomplishment and, um, that's great. Good.

Congratulations to you on that. But of course you, you know, you know what, that's a great opening, Brian. You're a distinguished engineer now.

How, how, how does one go about becoming a distinguished engineer? Give us a little bit of your story. Uh, so I've worked with Grafana for years.

It's, you know, I guess we're, we're releasing 10 anniversary, uh, I think I was realizing I've been working with it for seven now, um, and at Grafana for the last four years. Mm-hmm. Um, so I worked with Grafana as an open source contributor, uh, long before joining the company.

Mm-hmm. Mainly trying to build it so I could build applications with it myself to mm-hmm. At the time was working on mon monitoring, uh, hydroelectric turbine, so new kind of, uh, renewable energy systems, um, interesting.

And building observability systems for those. But, uh, turns out like the open source DevOps community has built better tools than our industrial, uh, counterparts. So really started working with Grafana then in building it into a system that allowed us to customize it for, for our needs.

Um, and then from there kind of got into working with, closely with Torco who, uh, started Grafana and been a close partner of his through the whole journey here. That's fantastic, man. That's interesting.

So you come really from kind of industrial engineering and, and, and that kind of holds space. Yeah, I mean, my background is slightly different than most of the people here who are very focused and we Grafana as a product focus on, uh, the observability space, people managing their, their clusters, their computer resources, et cetera. But as a product, we're across the board in sort of the, you build dashboards to understand how your applications, how your systems are working.

Uh, and I've sort of always pushed that we, we are, um, uh, we want to be opinionated but also agnostic, big tent, like where does your data come from? What are you trying to show with it? So Grafana as a product covers this range of being both open and extendable, but also trying to be a specific solution to solve real problems.

So, huge part of Grafana 10 has been trying to pull in what a bunch of those options are. Cuz it's, it's super powerful, but you've got so many knobs to turn and so many options to set up that it get getting started gets really hard. So a huge focus for this release has been on, uh, having better out of the box set up for the most standard deployments.

Sure. You know, look, I, I mean, I don't know, we sit on this side of the desk, right? I co-founded a few tech companies, software service SaaS.

Actually, we didn't call it SaaS then, but anyway, um, you know, and one of the observations I always found was, you know, 80% of our of our customers only use 20% of the features, right? And a lot of those knobs and dials just never got turned. They were left at the default.

And, you know, you put, and the, the thing about it is someone thought being able to turn that knob and do those things was a really important ability. And, but as I said, for 80% of the people it's not, and you know, it's, it's picking what's the right 20%, but you wanna have that ability when, when you do need to turn that up, right? And, and that that's what, you know, it's a little science, little art, but that's what makes it fun.

Ryan, I I want to, um, I mentioned the, the conferences going on. If you wouldn't mind, you've been there. Um, you're there.

Can you give people, give our audience maybe who aren't able to join in a, a kind of sense of what's happening? Yeah, it's, uh, I mean, I would say it's really a celebration in many ways of, uh, sort of recognizing how far we've come, uh, 10 years is, you know, round numbers. So we, um, will celebrate it both with the 10 year anniversary, with the new software release.

Um, some other big numbers is just recognizing that there's, we've got 20 million active kind of users all the time, which just, uh, still astonishes me, um, kind of how our, our sort of modest open source product has grown and so many people use it and enjoy using it. So, um, we also have about 2000 contributors, um, which is wow. Uh, also that should be based, um, uh, it's, it's a, as I was saying earlier, like it's a product that you can extend for your own needs that we're happy to connect to kind of any system producing data.

Uh, custom visualizations really gives people a, a chance to, to tweak it enough so that they can answer their own questions. Um, uh, with G Griffon 10, you know, there's been a lot that's just polished, I'd say, of kind of reducing some of our options, making that more, uh, clear. But there's also a few kind of big new features.

Um, uh, one of those I'd say is this public dashboards. So we, Grafana for a long time has made dashboards, but you're kind of in a trusted operator environment. Um, so those weren't things that you could just, you know, open those, open those up to the world to use.

Um, so in this release we now have features that allow you to publish a dashboard that you can safely share with a ton of people. Um, just nice. Um, and kind of the parallel world is we have a new thing for, for private data sources in our cloud environment.

So how can you connect to your on-prem data sources and show those in our cloud product? Excellent. Hey Brian, I, if you don't mind, you know, the world we I'm living in these days is, I can't take three steps without someone talking to me about ai.

Ai Yeah. Well look, I let you go a couple minutes before we brought it up, but, um, you know, in version 10 here, is there any kind, cuz cuz I mean, when we talk about graf and we talk about observability and, and you know, the big data aspect of all of this, obviously it lends itself to M L A I and that kind of Thing. Yeah.

So the, the recent hype that we are everyone's talking about is the large language models. Um, right. L many ways to think about those is like, um, uh, auto suggest kind of given given input.

Where, where do you go with that? Mm-hmm. We're definitely exploring how to make that useful in our cloud products.

We have some AI solutions that are looking for anomaly detection within your data sources. Um, not necessarily large language model based, but we, like everyone in this space are exploring how, how these tools could be useful and how to help move things along. Um, I think I'm, I'm slightly more skeptical than many of kind of, at what level do the query prompts become harder to use than, uh, you know, than writing a query.

Um, and at what level can we better kind of, uh, um, articulate what inputs are actually required for those to, to, to work out. But mm-hmm. Um, Agreed.

Um, so, you know, version 10 is shipping now. Yep. Is that correct?

Yep. What is, like, what's the upgrade path? Is it strictly just patching, you know, or Go Ahead, Jen, as an interesting kind of maturity level of a product, um, uh, what's perhaps the most interesting is how much more effort we've done is kind of making sure this is a seamless upgrade and, uh, uh, have spent a lot of time making sure that this can, can happen very transparently to the user.

So there's actually no, no real upgrade path aside from installing it. There's no, um, nothing changes. I guess couple, couple things like Angular, which we've tried to deprecate forever with our, uh, uh, plugin system, um, is now deprecated by default.

Um, so you have to enable it. Um, but say we've been running this in our cloud environment for a while, um, essentially the, the path forward on this one is really about us aiming for continuous deployment. Um, though we have these large round number releases, we're really aiming for continuous deployment of always adding new features, checking how they work over time.

Um, so there's always this conflict with the, the big, big number releases versus a cloud first strategy. So. Excellent.

Um, you know what we didn't mention, I mean, look, if people out here don't know the URL to go find out more about Grafana or Grafana Labs, which, which website's the best thing for them? Best place for 'em. com is certainly the, the best, uh, high level entry point to it.

Um, uh, it has, you know, the downloads, tutorials, et cetera. I mean, we, Grafana Labs, uh, I mean, today we're talking mainly about Grafana, the UI product that, uh, pulls together a whole bunch of sources. But as a company, we also work a lot with Prometheus, um, Loki and Tempo for a full observability stack.

Um, absolutely. Metrics, logs and traces all as a single, single pane of glass, single view. You're just different sources of data coming into a single pane, and that kinda is where the UI comes in.

Well, Ryan, listen, I want to congratulate you and the whole Grafana team on 10 years. That's an accomplishment, man, right? I, I, you know, we, uh, sometimes we, we, what is it that the days are long and the months and years are short, right?

So I know over 10 years, that's a lot of toil, a lot of work to become a distinguished engineer. And, and then you look back and it's like, my God, 10 years, it seems to gone really quickly, but you know, it, it's, it's day by day. So looking forward to the next 10 years, Yeah, it's exciting to, to work on software that people, you know, that help make people, that let people make good decisions and understand what's happening.

Um, and to, to like, to use it is a pretty Absolutely. All right. We're gonna let you get back to the conference.

Enjoy it en enjoy the moment and, and please keep us posted on progress at Grafana. Great. Thank you so much.

All right. Ryan McKinley, distinguished engineer at Grafana Labs here on Text Drug tv. We're gonna take a break and we'll be right back.