Connect with attendees on Slack in #grafanacon channel
About this session
Speakers
Speaker 1 (00:00): For those who don’t know, Grafana Labs really deliberately, strategically empowers engineers. Engineers get a lot of agency to push and implement stuff which they actually want to see. And we dog food all our offerings. And this is why personally I believe we have so good offerings and the quality of our software is so good. We really do use the stuff which we produce. We also have a system of hackathons where four times a year every employee at Grafana — Go-to-Market, Engineering, G&A — everything can take a whole week and work on whatever they like. The only thing which they have to do at the end is to present a five minute video and then we rank and we vote and we have winners and everything. Almost 50% of the hackathon projects have already shipped or are in the process of being shipped. And not everything ships, but a huge fraction does. And also a lot of the features, which you see, a lot of what you saw on stage, a lot of what you’re going to see on stage is actually based initially on hackathons, but not all of this is production grade. Some of this is just fun, like for example, Super Grot.
Speaker 2 (01:20): Hey everyone, my name is Gianni and I’m excited to be here to share something that we built during our most recent hackathon. So if you’ve ever been in a hackathon, the energy can be wild with people prototyping new features and pushing the boundaries of what is possible to do with our products. However, at the last hackathon, Domas and I decided to do something a little different, a little fun, a little tribute to a little yellow dinosaur, our mascot, Grot. So we built something that was fun for us, something a little bit weird. So instead of explaining it, I will actually show it to you. Yes.
Speaker 3 (03:50): All right, thank you. So it’s an awesome platform or video game starring Grafana’s most adorable mascot. We built it using Phaser JavaScript video game engine and the [inaudible] in the browser, you can play it on almost any device. And of course we instrumented it with Grafana Faro web SDK, so it’s reporting level completion times scores, player devs, coins, and other data. There’s a public dashboard, you can find it on play.grafana.org. And if you post a good time or a high score, it’ll show up there for everyone to see.
Speaker 2 (04:31): If you want to try it out, and you definitely should, you can scan the QR code and play it on your device, or you can come and find us at the science fair section on the second floor where you can actually play it in a little arcade cabinet that we have built for the conference. So please come check it out. We’ll see you there. And now we continue talking about birds, about dinosaurs, from dinosaurs to birds with the next project. Birds Drilldown. Thank you.
Speaker 4 (05:13): Yeah. Hi. So we are Ivana and Sven and we want to talk about our little fun hackathon project, but for now, let’s talk about Drilldown apps. As we’ve just heard in the keynote: Metrics, Logs, Traces, and Profiles Drilldown are now GA, which is really amazing. You don’t need to know LogQL, PromQL. TraceQL anymore to write your queries, Drilldown Apps really unlock a queryless and effortless, easy, opinionated solution to query your data and get insights into the data that you are looking for. And they’re really awesome. I can just encourage everyone to go and try them out. However, I think there’s something missing in this list. What do you think, Ivana?
Speaker 5 (05:55): I think you’re very right, Sven. There is something missing. Where is the Birds Drilldown? And if you’re all sitting here, very confused right now. No worries. Let me provide a little bit more context. Me and Sven, we have a little side hustle, side quest going on. We are on a mission to make Grafana recognized, not just as a monitoring company, but very specifically, a birdsong monitoring company. So when someone says Grafana, we want people to immediately think, “Oh yeah, dashboards and birds.” It started all right here at the last year’s GrafanaCON where we shared how easy it is to create a bird monitoring solution. If you weren’t here or if you’ve missed it, here is a super quick masterclass. All you need is Raspberry Pi. Then you need microphone, one hour of the spare time. And we have written a very detailed blog post that you can just follow.
(06:56): Of course, you’ll need some birds that you will monitor. Inspired by Drilldown Apps and powered by awesome open source technology such as Grafana Scenes, Grafana App platform and Grafana LLM. We’ve built a full-blown exploratory experience for birds. With Birds Drilldown, you don’t need to write queries anymore or build dashboards anymore. All you need is to open the app and start exploring. You will immediately see the latest bird song detection. You will be able to spot interesting patterns and learn how things such as weather or air quality influences the quantity of the bird songs. Moreover, if you are interested to learn more about which bird species sing the most or which bird species songs vary the most, you can simply sort by that and then align this with environmental data to learn some interesting patterns.
Speaker 4 (08:01): So every bird has its own page in the app where you can see a picture of a bird. You can post some descriptions or you can read about it. You can even listen to birds so you don’t have to go outside anymore, I guess. And nowadays, no great project comes without AI. So we have a 24/7 AI bird expert directly built into the app using the Grafana LLM plugin. And it’s always there to answer your questions like pronunciation of a penguin or it’s pooping schedule, for example. So at our science fair, we showcase three main parts of our app. First of all, the bird monitoring setup, so the hardware, Raspberry Pi, microphone, and so on. We wanted to bring real birds, but that idea flew out of out of the windows. So we brought bird plushies that make real bird sounds, and obviously we demoed our app with Drilldown. So everything is hands-on. See you at the science fair. And if you’re not that much into birds, but more into games, back to Domas and Bogdan.
Speaker 6 (09:11): So hello everybody. I’m Bogdan and I’m here with my colleague Domas. We are both software engineers and we are going to start by mentioning our first hackathon project. That would be Faro. Some time passed, and then we had our second hackathon. And during the last day of registrations, we are there asking, what should we do? It should be meaningful, it should be fun, it should be useful, it should be a normal hackathon project. But we had literally no idea. So we wanted to answer a question. Does Grafana run Doom? So without any further…
Speaker 7 (09:56): Running Doom in virtually every device that can display pixels has always been a thing. MacBook Touchbars, ATMs or even pregnancy tests. But what about Grafana? Can Grafana run Doom?
Speaker 3 (11:02): So we created a data source plugin for Doom. Thank you. It runs Websocket’s doom, which is a fork of Chocolate Doom compiled to Web Assembly by CloudFlare. It’s a rendering video into WebGL Buffer in memory, which we then translate to Grafana data frames with 256 fields, one per color. And then we stream it to a regular time series panel with 256 overrides. So for every frame of Doom, it’s rendering 64,000 data points. And on my machine, I got it to run at 15 frames per second.
Speaker 6 (11:55): No, that was my machine. Your machine did not run it to 15. Come on.
Speaker 3 (12:03): All right. We learned a few things along the way on how to max out performance for live streaming time series panels. It’s important to avoid expensive on-the-fly calculations by explicitly setting minimum and maximum values for your data on the data frames. Also, to keep the list of fields constant, even if they don’t always have any values for every frame. And push new data frames on window request animation frame tick. Also, if running Chrome, it really helps to enable Canvas out of process restorization flag, if it’s not on already. So it is live on play.grafana.org, you can find it there or just scan the QR code or please come and see it in the science fair and talk to us. Thank you.
Speaker 6 (12:43): Thank you.