Netdata

Netdata develops an open-source, real-time infrastructure monitoring and troubleshooting platform. Its architecture has three parts: the Netdata Agent, a core monitoring engine licensed under GPL v3+ that handles collection, storage, machine learning, alerting and exports and runs on servers, cloud instances, Kubernetes and IoT devices; Netdata Cloud, a closed-source layer providing user management, RBAC, horizontal scaling and centralized alerting without centralizing metric storage; and Netdata UI, the dashboard and visualization layer distributed under the NCUL1 license and included in standard packages.

The product collects per-second metrics with automatic discovery and no required configuration, and the company states it ships more than 800 collectors and notification integrations. Agents can stream to centralized "Parent" nodes for aggregation, long-term retention, high availability clustering and replication. Netdata reports an agent footprint of roughly 5% CPU, 150MB RAM and 3GB of disk for more than a year of retention, and about 0.5 bytes per sample using tiered storage. Coverage spans Linux, FreeBSD, macOS and Windows and includes system resources, storage, network interfaces and protocols, hardware sensors, processes, logs (systemd-journal, Windows Event Log, ETW), containers, VMs, synthetic checks, packaged applications, cloud provider infrastructure and custom applications via OpenMetrics and StatsD, with OpenTelemetry noted as forthcoming. Additional capabilities marketed on the website include unsupervised machine learning for anomaly detection, an "Anomaly Advisor" and AI co-engineer features, Model Context Protocol support for connecting AI tools to observability data, eBPF monitoring, network topology mapping, NetFlow traffic analysis and SNMP device and trap monitoring, plus native iOS and Android apps.

Deployment options range from the free open-source agent to a managed SaaS platform and an on-premises/air-gapped installation of Netdata Cloud aimed at data-sovereignty and compliance requirements. The company states SOC 2 Type 2 certification and that metrics remain on customer infrastructure with only views or metadata reaching the cloud.

Founding story

Costa Tsaousis was an executive (described as COO in the project README and as a C-level executive in Greece in press coverage) at a company where, in 2013-2014, a significant share of cloud-based transactions failed silently and existing troubleshooting tools could not identify the root cause. Frustrated that monitoring systems offered few metrics at low resolution, scaled poorly and were costly, he began writing his own tool at nights and on weekends despite not having coded much in years. He spent about two years building it and released it as open source in 2016, where it went viral after a single Reddit post, gaining roughly 10,000 GitHub stars in a week.

Business model

Netdata combines a free, open-source agent (GPL v3+) with commercial cloud offerings. At the time of the 2020 funding announcement, both the open-source Netdata Agent and the closed-source Netdata Cloud were offered free of charge, with a stated intent to deliver future commercial products. The current website advertises pricing plans, enterprise features such as SSO and RBAC, an on-premises Netdata Cloud deployment sold through a sales motion, and a free community tier of Netdata Cloud.

Paid subscriptions to Netdata Cloud alongside a free tier; in 2019 the founder described the paid SaaS tier as charging per seat rather than by the amount of infrastructure monitored. Enterprise/on-premises deployments are sold via direct contact with sales.

Traction

The GitHub repository shows roughly 80,200 stars, 6,600 forks and 23,131 commits, while the website cites 76k GitHub stars and more than 668 million cumulative Docker pulls. Earlier milestones reported in September 2020 included more than 3 million users worldwide, over 5,000 new users per day, nearly 50,000 GitHub stars and close to one million Docker Hub downloads per day; in 2019 the founder cited about half a million downloads per day. Named users referenced on the website include Codyas and LANCOM Systems.

Latest developments

The current website promotes autonomous, AI-native monitoring with built-in self-learning models, an AI co-engineer, anomaly advisor, root cause and blast radius analysis, and Model Context Protocol integration for connecting AI assistants to observability data. Other recent additions include zero-pipeline log management, OpenTelemetry, unified network monitoring (live topology viewer, NetFlow traffic analyzer, SNMP device and trap monitoring, device auto-discovery), iOS and Android apps, an on-premises deployment of Netdata Cloud with air-gapped support, SOC 2 Type 2 certification, and a published roadmap covering AI-native observability, full-stack signal coverage, operational intelligence and enterprise platform maturity, with agent releases every six weeks and continuous delivery for the cloud. A webinar on fleet monitoring was scheduled for August 26.