See Every Second of Your Bare Metal Performance

Yes. Netdata provides comprehensive monitoring for 100% of Hetzner servers - DELL, PX series, EX, AX, and all models. For servers with IPMI (DELL, PX60/70/90/120), we monitor via BMC. For the 70% without IPMI (post-2019 policy), we use lm-sensors, SMART, and system metrics. You get complete hardware health monitoring regardless of model.

Netdata provides 30+ notification channels that don’t require email: webhooks (Discord, Slack, Microsoft Teams), incident management (PagerDuty, Opsgenie), mobile/chat (Telegram, Pushover, ntfy), and SMS (Twilio, MessageBird). These integrate in minutes and work immediately - no SMTP relay configuration required. Email is optional, not mandatory.

Hetzner SysMon provides basic availability checks (ping, HTTP, ports) with 6-minute minimum alert delay and email-only notifications. Netdata provides comprehensive infrastructure monitoring: CPU, RAM, disk, network, processes, hardware health, logs, and 800+ application integrations. Alerts trigger in sub-2 seconds with 30+ notification channels. SysMon is free but limited; Netdata offers predictable per-node pricing with complete observability.

60 seconds per server with one-line install. Auto-discovery finds all services, containers, and applications immediately. Dashboards generate automatically. 400+ alerts activate instantly. For 100 servers: Day 1 (install all agents), Day 2-3 (optional Parent setup for centralization), Day 3-7 (optional Cloud integration). Total time to production: 1-7 days vs 2-4 weeks for Prometheus/Grafana.

Yes, both. Netdata monitors Hetzner Cloud instances, dedicated bare metal servers, and hybrid deployments. For Cloud, we provide Hetzner service discovery integration. For dedicated servers, we offer comprehensive hardware monitoring (IPMI, lm-sensors, SMART). Same agent works across all Hetzner infrastructure types.

Netdata Business: volume discounts apply for 100 nodes. This includes unlimited metrics, logs, users, containers, ML anomaly detection, AI troubleshooting, and all features. Compare to commercial alternatives for significant cost reduction. Open Source version is free for unlimited servers with community support.

Multi-layer approach: (1) IPMI via freeipmi for DELL/PX series with BMC, (2) lm-sensors for CPU/motherboard temperature on all servers, (3) SMART monitoring for disk health on all drives, (4) NVMe-specific metrics for modern SSDs, (5) GPU monitoring for NVIDIA/Intel/AMD. This provides comprehensive hardware visibility on 100% of Hetzner servers, not just the 30% with IPMI.

Yes. Netdata provides everything Prometheus+Grafana offers plus more: (1) Zero configuration vs 2-4 hours setup, (2) Auto-generated dashboards vs manual building, (3) No query language required vs PromQL, (4) Built-in ML anomaly detection vs none, (5) 97% less disk I/O, (6) 88% less RAM, (7) 16× faster queries. Netdata is Prometheus-compatible (scrapes OpenMetrics, exports to Prometheus) so you can migrate gradually.

Yes. Netdata Agent works completely offline with local dashboards. For centralized dashboards in air-gapped environments, deploy Netdata Parents (same software, configured as aggregation points). For complete control plane, Netdata Cloud On-Premises runs entirely in your datacenter. All observability data stays on your infrastructure - only metadata travels to Cloud (optional).

18 unsupervised k-means models train per metric using 6-hour windows with 3-hour staggered intervals. Anomaly flagged only when ALL 18 models agree (consensus-based detection). This achieves 99% false positive reduction with 10^-36 theoretical false positive rate. Models train automatically in 15 minutes, retrain every 3 hours, and run as low-priority background tasks using ~18 KB RAM per metric. No configuration required - works on all metrics from day one.

Netdata continues monitoring locally. Agent stores data on disk (multi-tier retention: 14 days + 3 months + 2 years default), evaluates alerts, and serves local dashboard at http://localhost:19999. When connectivity restores, Agent automatically streams buffered data to Parents/Cloud with zero data loss. This ensures monitoring works during network partitions, Hetzner outages, or maintenance windows.

Yes. Netdata monitors Storage Boxes via SNMP (for network-attached storage) or by mounting them and monitoring as filesystems. We track disk usage, I/O operations, network throughput, and access patterns. For Backup Space, monitor via mounted filesystem metrics. Custom collectors available for specialized Storage Box monitoring needs.

Netdata’s agent-based architecture eliminates API dependency. Monitoring runs locally on each server - no continuous API polling required. For Hetzner Cloud service discovery (optional), we use API token for initial discovery only, not ongoing metrics collection. This avoids the 3,600 requests/hour limit that affects cloud-native monitoring platforms. You can monitor unlimited servers without API constraints.

Yes. Netdata works on public IPs, private IPs, vSwitches, VPNs, and any network topology. Agent→Parent streaming uses TLS encryption (equivalent to VPN security). We’re not constrained by Hetzner’s 100-server private network limit - you can monitor 500+ nodes per Parent cluster across any network configuration. Optional: use private networks for efficiency, but not required.

Everything: unlimited metrics (no per-metric charges), unlimited logs (no per-GB charges), unlimited users (no per-seat charges), unlimited containers (no per-container charges), ML anomaly detection (18 models per metric), AI troubleshooting (Chat + Insights), 400+ pre-configured alerts, 30+ notification channels, REST API access, and all features. Only active nodes count (P90 billing excludes spikes). No hidden costs, no data volume penalties.