 Live from San Diego, California, it's theCUBE. Covering KubeCon and CloudNativeCon. Brought to you by Red Hat, the CloudNative Computing Foundation and its ecosystem partners. Welcome back to theCUBE. I'm Stu Miniman, my co-host is John Troyer and you're watching theCUBE here at KubeCon, CloudNativeCon 2019 in beautiful and sunny San Diego today. Happy to welcome to the program a first-time guest, Tom Wilkie, who's vice president of product at Grafana Labs. Thank you very much. Thank you so much for joining us. Thank you. All right, so Grafana, it's on your t-shirt. We've been hearing customers talking about it and the like, but why don't you introduce the company to our audience and where you fit in this broad landscape here at the CNTF show. Thank you, yeah, so Grafana is probably the most popular open source project for dashboarding and visualization. Started off focused on time series data, on metrics, but really recently it's branched out into log analysis and tracing and all of the kind of aspects of your observability stack. All right, so really big, you know, broad topic there. We know many of the companies in the space. There's been many acquisitions, you know, recently in this. Where do you fit in the ecosystem? I saw like databases, like a big focus when I look at the company website. Bring us inside a little bit of the products we're offering the customers. Most vendors in this space will sell you a monitoring product that includes the time series database, normally includes visualization and some agent as well. Grafana Labs and Grafana open source projects very focused on the visualization aspects. So we are data source agnostic and we have backends for more than 60 different data sources. So if you want to bring together data from, let's say, Datadog and combine it with some open source monitoring from Prometheus, you can do that with Grafana. You can have the dashboards and the individual panels in that dashboard combine data from multiple different data sources. And we're pretty much the only game in town for that. You can think of it like Tableau allows you to plug into a whole bunch of different databases for your BI with that, but for monitoring and for metrics. Well, so Tom, maybe let's, before we get into the products and more of the service and the conference here, let's talk, well, on the front page of your website you use the O11Y word. So we've said words like monitoring here. We use words like management. We use words like ops. Observability is a hot topic in the space and for people in the space, they had some nuances. And so can you just maybe let the viewers and us know a little bit about how the space is looking at this and how you all feel about observability and what everybody here who's running some cloud native apps needs to actually function in production. Yeah, so I think you can't talk about observability without either being pro or for the three pillars, right? So people talk about metrics, logs and traces. I think what people miss here is that it's more about the experience for the developer. Grafana and what we're trying to achieve is all about giving engineers and developers the tools they need to understand what their applications and their infrastructure are doing. So we're not actually particularly picky about which pillars you use and which products you use to implement those pillars. But what we want to do is provide you with an experience that allows you to bring it all into a single user interface and allows you to seamlessly move between the different sources of data and hopefully combine them in your analysis and in your root cause of any particular incident. And that for me is what observability means. It's about helping you understand the behavior of your application. In particular, I mean I'm a software engineer by trade. I'm still on call. I still get paged at 3 a.m. occasionally. And having the right tools at 3 a.m. to allow me to as quickly as possible figure out what happened and then dive into a fix, that's what we're about at Grafana Labs. All right, so Tom, one of the things we always need to understand and show here, there's the project and there's the company. Yep. Help us just kind of understand the difference the products, the mission of the company and how that fits with the project. So the Grafana project predates the company and it was started by Tawkel. He, you know, he saw a spot for like needing a much better kind of graphical editing of dashboards and making the kind of metrics way more accessible to your average human. Grafana Labs started really to focus on the IT and monitoring observability use cases of Grafana. But the project itself is much broader than that. We see a lot of use cases in industrial, in IoT, even in BI as well. But Grafana Labs as a company, we're focused on the monitoring side of things. We're focused on the observability. So we also offer, we, like most companies, we have an enterprise version of Grafana. It has a few data sources for commercial vendors. So if you want to get your data dog or your Splunk into Grafana, then there's a commercial option for that. But we also offer a hosted observability platform called Grafana Cloud. And this is where we take the best open source projects, the best tools that we think you need as an engineer to understand your applications and we host them for you. We operate them for you. We scale them, we upgrade them, we fix bugs. We, so Grafana Cloud's predominantly a hosted Prometheus, a hosted Graphite, and a hosted Loki, our log aggregation system. All combined and brought together with the Grafana front-end. So yeah, like two products, a bunch of open source projects. Grafana Labs employs four of the Prometheus maintainers. I mean, I'm one of the Prometheus maintainers. We employ Graphite maintainers, obviously a lot of Grafana maintainers, but also Loki, I'm trying to think like, there's just so many open source projects we get involved with that really it's about synthesizing an observability platform out of those. And that's what we offer as a product. So you've recently had an announcement that Loki is now GA, can you talk just a little bit about Loki and aggregation and logs and what Loki does? Yeah, I'd love to, yeah. A year ago in Seattle actually, we announced the Loki project. It was super early, I mean, I just basically been finishing the code on the plane over. And we announced it and no one, I think could have predicted the response we had. Everyone was so keen and so hungry for alternative to traditional log aggregation systems. So it's been a year and we've learned a hell of a lot. We've had so much feedback from the community. We've built a whole team internally around Loki. We now offer a hosted version of it and we've been running it in production now for over a year, doing some really great scale on it. And we think it's ready for other people to do the same. One of the things we hear, especially at shows like this is I really, developers and the grassroots adopters come to us and say, we really love Loki. We really love what you're doing with it. But my boss won't let me use it until it goes V1. And so really yesterday we announced it's gone V1. We think it's stable. We're not going to change any of the APIs on you. We would love you to use it and put it into production. All right, we'd like to hear a little bit more about the business side of things. So I believe there's some news around funding. How many people you have, how many, can you parse for us how many customers of the projects versus how many customers of the company's products? Well, we don't call them customers of the projects. They're users. Yes, yes, of course. But I'm from a company where we have hundreds of customers. I don't believe we make our revenue figures public. So I'm probably not going to dive into them. But I know the CEO stands up at our yearly conference and discloses what our revenue of the last year was. So I'll refer you to that. The funding announcement, that was about a month ago. We raised a great round from Lightspeed, 24 million, I believe. And we're going to use that to really invest in the community, really invest in our projects and build a bit more of a commercial function. The company is now about 110 people, I think. It's growing so quickly. I joined 18 months ago and we were 30 people. And so we've almost quadrupled in size in the last year and a half. So keeping up is quite a challenge. The two products I've already touched on, a few hundred customers. And I think we're really happy with the growth. We've never had any institutional funding before this. The company's about five years old. So we've been growing based on organic revenue and barely profitable, but reinvesting that into the company. And it's growing really well. Also one of the, I mean, it's not that unique, I guess, but we're remote first. We have more than 50% of our employees work from home. I work from my basement in London. We have a few tiny offices, one in Stockholm and one in New York, but we're really keen to hire the best people wherever they are. And we invest a lot in travel. We invest a lot in the right tools and getting the whole company together to really make that work. It's a really fun place to work. Well, Tom, we're still in the business here and I don't know how much time you've spent at the booth this year, but can you compare? I mean, we've been talking about the growth of this community and the growth of this conference. Can you compare, say, this year to last year, the people coming up, their maturity, the maturity of their production, et cetera. Are they ready to buy? Are they still wondering what this Kubernetes-y thing is? Where is everybody this year and how does that, how has it changed? Yeah, and that's a good question. We're definitely seeing people with a lot more sophisticated questions. The conversations we're having at the booth are a lot longer than they've been in previous years. The, you know, in particular, people now know what Loki is. We only announced it a year ago and a lot of people asking us very detailed questions about what scale they can run it at. Otherwise, yeah, I think there is starting to be a bit more commercial intent at the conference. So a few more buying decisions being made here. It's still predominantly a community-oriented conference and I think the, I don't want that to go away. Like that's one of the things that makes it attractive to me and I bring my whole team here and that's one of the things that makes it attractive to them. But there is a little bit more sales activity going on, for sure. Any updates to the tracing and monitoring observability story of the projects here at CNCF this year, since you're part of the Prometheus project? Yeah, so we actually, we had the Prometheus conference in Munich two weeks ago and after each Prometheus conference, the maintainers like to get together and kind of plan out the next six months of the project. So we started to talk about adding support for things like exemplars into Prometheus. This is where for each histogram bucket, you can associate an example trace that goes, that contributed towards that histogram, that latency. And then you can build nice user interfaces around that so you can very quickly move from a latency graph to example traces that cause that. So that's one of the things we're looking to do in Prometheus. Of course, Jager graduated just a week ago, I think. We're big users of Jager internally at Grafana Labs. And actually on our booth right now, we're showing a demo of how we're integrating visualization of distributed tracing into Grafana. So you can, using the same approach we do with metrics where we support multiple backends, we're going to support Jager, we're going to support Zipkin, we're going to support as many open source tracing projects as we can with the Grafana UI experience and being able to seamlessly kind of switch between different data sources, metrics, all the way to logs, all the way to traces, within one UI and without ever having to copy and paste your query and make mistakes and kind of translate it in your head. All right, Tom, give us a little bit, look forward. A lot of activities, things going, graduating and pulling things together. So what should your users be looking for kind of over the next six to 12 months? That's a great question, yeah. I think we do a yearly release cycle for Grafana. So the next one we're aiming towards is Grafana 7. Like for me, Grafana 7 is going to be all about tracing. So I really want to see the demo we're doing. I want to see that turned into like production ready code, support for multiple different data sources, support for things like exemplars, which we're not showing yet. I want to see all of that done in Grafana in the next year. We've also massively been fleshing out the logging story. With Loki, we've been adding support for extracting metrics from the logs. And I really think that's kind of where we're going to drive Loki forward in the future. And that really helps with systems that aren't really exposing metrics like legacy systems where the only kind of output you get from them is the logs. Beyond that, yeah, I mean, the worlds are kind of oyster, I think. I'm really keen to see the development of open telemetry and we're just starting to get involved to that project ourselves. I'm really interested to kind of talk to people about what they need out of a tracing system. We see people asking for a hosted tracing systems, but IMO is very much like pick the best open source ones. I don't think that's emerged yet. I don't think people know which is the best one yet. So we're going to get involved in all of them, see which ones are community kind of coalesces around and maybe start offering a hosted version of that. Yeah, final thing is, what advice do you have for users? Obviously, you like the open source thing, but they're hearing about observability everywhere. There are the whole APM market is moving this direction. There's acquisitions as we talked about earlier. There's so many moving pieces and a lot of different viewpoints out there. So just from a user, how will things, what makes their lives easier and what advice would you give them? Yeah, no, definitely. I think a lot of vendors will tell you like to pick a vendor who's going to help you with this journey. I would say like pick a vendor you trust who can help you make those decisions. Like find someone impartial who's going to not make, not try and persuade you to buy their product. So we would, I would encourage you to try things out to dog food and to really like invest in experimentation. Like there's a lot going on in the observability world and in the cloud native world and you've got to try it and see what fits. Like we embrace this composability of the observability ecosystem. So like try and find which choices work best for you. Like whenever I talk to them. You still have to lick all the cupcakes in 2019. I think, I mean, I would, it depends on your level of kind of maturity, right? And sophistication. I think if this is really important to you, you should go down that approach. You should try them all. If this is not one of your core competencies, then maybe going with a vendor that helps you is a better approach. But I come from the open source world and I like to see the whole ecosystem and all the different players and all the different new and exciting ways to solve these problems. So I'm always going to encourage people to have a play and try things out. All right. Tom, final word, Loki. Explain to us when you're coming up with it, how you ended up. Are you the God of Mischief? Well, so the official line is that Loki is the North mythology equivalent of Prometheus in Greek mythology. And Loki, our logging project, is Prometheus inspired logging. So we've tried to take the operational model from Prometheus, the query language from Prometheus and the kind of cost efficiency from Prometheus and apply it to logs. But I will admit to being a big fan of the Marvel movies. All right. Tom Wolke, thank you so much for sharing the update time for Fana Labs. Definitely look forward to hearing updates from you and the future. Thank you. All right. For John Troyer, I'm Stu Miniman. Back with more coverage here from San Diego. Thank you for watching, thank you for watching theCUBE.