Netdata, the company behind the eponymous open-source, distributed, real-time, performance, and health monitoring solution for systems and applications, has announced it completed a new round of financing. Totaling $14.2 million and led by Bessemer Venture Partners, with participation from existing investors Bain Capital Ventures and Uncorrelated Ventures, it brings Netdata's total raised in Series A to $31 million.
"Giving SREs, DevOps engineers, sysadmins, and developers a way to gain visibility into their infrastructure in minutes, with zero configuration, thousands of metrics, and milliseconds from data collection to visualization."
Interestingly, Netdata does this without having a single paid service or a single paying customer so far. ZDNet connected with Tsaousis to get the whole story.
Open-source application monitoring with a twist
You have a pet peeve that you think nobody has adequately addressed before. So, you bite the bullet and start building your own software solution to address it. It works, and you want to share with the world, so you open source it. People like it, use it, and contribute to it, so it gets traction. The potential is there, so you want to scale it up and build a business around it.
You keep the core offering open source and then take it, perhaps extend it with value-add enterprise features, and offer it as a managed service in your own cloud. You have to convert part of your user base to paying customers. Clients who need the features and don't want to run their own service pay for it.
You got an open-source business and community going. Of course, you have to keep developing your offer and worry about differentiating from competitors. If you are successful enough to draw attention from cloud vendors, your licensing and your features must make it hard for them to compete against you with your own product.
That's how the open-source business playbook looks like these days. In many ways, Netdata plays by that book. There are some interesting ways in which it deviates from the norm, such as offering everything for free, and not worrying about cloud vendors stealing its lunch. But let's take it from the start. Tsaousis said he started Netdata in 2014 out of frustration:
"I was really pissed off with the monitoring systems that existed. I couldn't believe that most of the monitoring systems even today provide just a plain view of the infrastructure. So they cherry pick just a few important metrics, very limited information. And they try to use this information to provide a monitoring solution.
The problem with this, however, is that it is too abstract. Many insights are lost for the sake of scalability, and we have very limited monitoring solutions. I decided to to build Netdata in a completely different way with completely different principle."
The key architectural decision on which Netdata's approach is based is distribution. Netdata monitoring agents are deployed locally, and the data they collect is stored there, not retained centrally. Netdata server only gets metadata, not data, metrics are fine-grained (at 1-second granularity), and calculations are done on the fly.
This is an interesting choice, but isn't there a tradeoff in terms of execution speed, if everything has to be fetched on the fly? Netdata agents were harder to implement, as they utilize parallelization, but their performance is superb, said Tsaousis.
Agents for nothing, cloud service for free
In fact, Tsaousis went on to add, this is precisely what enables Netdata to offer its product, even on the recently unveiled Netdata Cloud, for free. Νot storing data can lower the bills significantly. Still, how are compute costs paid? Where do the 50 people in Netdata's (remote) team get their salaries from? And don't Netdata's investors expect a return on their investment? Could it be another case of "when the service is free, the product is you?"
Tsaousis was reassuring. There will be paid services, he said, but not yet. There is a catch, but it's clear: Today, most of Netdata's features are limited to a single agent. Going forward, they will all be kept free, but infrastructure-wide features will be added, and users will have to pay for those. In other words, Netdata is building its user base, with the aim of converting part of it to paying customers later.
Tsaousis, who identified as an open-source aficionado, referred to examples such as Github or Slack as inspiration for the business model that Netdata wants to pursue. Base service for free, value-add service as a paid tier. And what about the competition, especially from cloud vendors?
Another aspect that Netdata emphasizes is usability, mostly in terms of out-of-the-box functionality. Netdata agents are built with sane defaults and with the goal of auto-detecting as much as possible with as little intervention as possible, although customization is possible too, said Tsaousis:
"Observability is about three pillars: metrics, logs, and traces. Netdata only addresses metrics today. We tried to simplify the lives of developers as much as possible. Netdata is extremely configurable. But I think that what makes makes Netdata so good is that it is so easy and it has such sane defaults when you use it and when you first install it."
In terms of competition, Netdata is not out to displace solutions like Prometheus, Grafana, Elastic, or Splunk. Seventy-five percent of Netdata's user base uses it alongside other solutions, Tsaousis said. There are many solutions used for monitoring, he went on to add, but what the world lacks is a troubleshooting platform.
Tsaousis said the new investment will further Netdata's vision by enabling them to build upon their community momentum to deliver innovative solutions in both the open-source project and future commercial products.
Netdata has also hired new executives (CMO and CFO) recently and seems to be growing in a healthy way, considering it started as a company in 2018. What remains to be seen is whether this growth can actually lead to sustainability and profitability.