Access Everything From Anywhere While Data Stays Home

Netdata Cloud stores zero metric or log data. All observability data stays on your Agents and Parents. Cloud acts as an intelligent control plane that orchestrates queries to your infrastructure and aggregates results in real-time. Only metadata (node names, chart definitions, alert configurations) is synchronized to Cloud. This architecture satisfies GDPR, HIPAA, and PCI DSS requirements by design while enabling infrastructure-wide visibility.

P90 billing uses double 90th percentile methodology. Daily P90 excludes the top 10% of each day’s usage (brief spikes don’t inflate bills). Monthly P90 excludes your top 3 usage days per month. Example - if you normally run 50 nodes but spike to 200 for load testing and 300 during an incident, you’re billed for 50 nodes, not 300. This protects against temporary scaling, testing, and incident response costs.

Yes, through Netdata Cloud On-Premises. This provides the full Cloud control plane hosted in your datacenter with complete feature parity. For partially restricted environments, Agents can operate in standalone mode with local dashboards, then connect to Cloud when network access is available. All data remains on-premises regardless of deployment model.

Netdata Spaces provide physical isolation - separate infrastructure, billing, user databases, and RBAC policies per customer. Rooms within each space enable logical organization (production, staging, services). A single admin interface manages all customer spaces. Customers cannot see each other’s data. Each space has independent billing with transparent per-node pricing.

All your observability data remains on your infrastructure (Agents and Parents) because Cloud never stored it. The open source Agents continue working with local dashboards. You can export configurations via REST API. Switch to standalone mode or Enterprise On-Premise without data migration. Zero vendor lock-in by design.

Distributed architecture eliminates central bottlenecks. Each Agent processes data locally. Cloud distributes queries in parallel to Agents and Parents, aggregating results. Adding nodes doesn’t affect existing performance because work stays distributed. Proven at massive scale processing 4.5+ billion metrics per second globally with sub-2-second query latency.

Netdata AI is significantly more effective because it has access to (1) real-time ML-based anomaly detection on ALL metrics, (2) per-second high-resolution data, (3) bottom-up component-level alerts, and (4) anomaly enrichment in every query. This enables accurate root cause analysis that generic AI tools cannot achieve. Includes AI Insights (automated reports), AI Troubleshoot (interactive investigation), and AI Chat (natural language queries). Business plan includes 10 free sessions per month.

Netdata Cloud provides zero-configuration algorithmic dashboards vs Grafana’s manual dashboard building. No query language required (NIDL point-and-click vs PromQL). Built-in ML on every metric vs optional add-ons. Data sovereignty by design vs centralized storage. Both support enterprise SSO and RBAC. Netdata focuses on operational simplicity, Grafana on visualization flexibility.

Netdata is SOC 2 Type 2 certified with independent third-party audit. Architecture satisfies GDPR (data residency by design), HIPAA (BAA available), PCI DSS (data sovereignty), and CCPA requirements. Comprehensive audit logs track all user actions and configuration changes. Enterprise SSO via OIDC, SCIM 2.0 for automated provisioning, LDAP/AD integration for role mapping.

Yes. Centralized alert notifications support PagerDuty, Slack, Teams, Discord, email, and custom webhooks. Agents can also send notifications directly (20+ integrations). For visualization, use the Grafana datasource plugin to query Netdata from Grafana dashboards. Export metrics to Prometheus, InfluxDB, TimescaleDB for data warehouse integration. OpenTelemetry support enables broad ecosystem compatibility.

Netdata Cloud SaaS is hosted by Netdata with automatic updates, managed infrastructure, and simplified operations. Enterprise On-Premises provides the same control plane hosted in your datacenter for air-gapped environments or complete infrastructure control. Both offer identical features (multi-tenancy, RBAC, AI, dashboards). On-Premises requires Kubernetes cluster and self-management. Choose based on connectivity requirements and operational preferences.

Deployment is fast. Agent installation - 60 seconds per node with one-line command. Auto-discovery - immediate for systems, containers, applications. First dashboards - 60 seconds after installation. ML anomaly detection - 15 minutes to first results. Full ML training - 2 days for all 18 models per metric. Enterprise features - hours for SSO, SCIM, and RBAC configuration. Most organizations achieve full production deployment in days, not months.

Support tiers include Community plan - public forums and tickets. Business plan - email/ticket support during business hours with SLA. Enterprise - 24/7 support with dedicated team, phone support, custom SLAs. All plans include comprehensive documentation, video tutorials, and active Discord community. Professional services available for implementation assistance, architecture design, and migration support.

P90 billing methodology protects against auto-scaling cost spikes. You’re charged based on sustained usage, not temporary peaks. Daily P90 excludes brief scaling events. Monthly P90 excludes your top 3 usage days. Example - if you normally run 100 nodes but auto-scale to 500 during traffic spikes, you’re billed for 100 nodes. This makes costs predictable despite infrastructure volatility.