Netdata
Netdata is a real-time, zero-configuration infrastructure monitoring platform that collects per-second metrics and uses built-in machine learning to detect anomalies without a query language.
Netdata auto-discovers what is running on a node and starts collecting per-second metrics immediately, training multiple unsupervised anomaly-detection models per metric at the edge rather than centralizing raw data. Storage is compressed to roughly half a byte per sample with tiered retention, and a parent-child architecture lets it scale horizontally to multi-million-sample-per-second workloads while keeping troubleshooting on interactive, query-language-free dashboards.
An independent university study is cited as finding it the most energy-efficient tool for monitoring Docker-based systems among those compared, alongside strong CPU, memory and execution-time results. It integrates with Prometheus, Grafana and common databases, and is commonly deployed via Docker or Kubernetes.
The project is distributed under GPL-3.0 and written in Go.