| 1 | # CPU Utilization |
| 2 | |
| 3 | Netdata's CPU usage depends on the features you enable. For details, see [resource utilization](/docs/netdata-agent/sizing-netdata-agents/README.md). |
| 4 | |
| 5 | ## Children |
| 6 | |
| 7 | With default settings on Children, CPU utilization typically falls within the range of 1% to 5% of a single core. This includes the combined resource usage of: |
| 8 | |
| 9 | - Three Database Tiers for storage |
| 10 | - ML for Anomaly Detection |
| 11 | - Per-second data collection |
| 12 | - Alerts |
| 13 | - Streaming to a [Parent Agent](/docs/observability-centralization-points/metrics-centralization-points/README.md) |
| 14 | |
| 15 | ## Parents |
| 16 | |
| 17 | For Parents, we estimate the following CPU utilization: |
| 18 | |
| 19 | | Feature | Depends On | Expected Utilization (CPU cores per million) | Key Reasons | |
| 20 | |:--------------------:|:---------------------------------------------------:|:--------------------------------------------:|:------------------------------------------------------------------------:| |
| 21 | | Metrics Ingest | Number of samples received per second | 2 | Decompress and decode received messages, update database | |
| 22 | | Metrics re-streaming | Number of samples resent per second | 2 | Encode and compress messages towards another Parent | |
| 23 | | Machine Learning | Number of unique time-series concurrently collected | 2 | Train machine learning models, query existing models to detect anomalies | |
| 24 | |
| 25 | To ensure optimal performance, keep total CPU utilization below 60% when the Parent is actively processing metrics, training models, and running health checks. |
| 26 | |
| 27 | ## Increased CPU consumption on Parent startup |
| 28 | |
| 29 | When a Parent starts up, it undergoes a series of initialization tasks that can temporarily increase CPU, network, and disk I/O usage: |
| 30 | |
| 31 | 1. **Backfilling Higher Tiers**: The Parent calculates aggregated metrics for missing data points, ensuring consistency across different time resolutions. |
| 32 | 2. **Metadata Synchronization**: The Parent and Children exchange metadata information about collected metrics. |
| 33 | 3. **Data Replication**: Missing data is transferred from Children to the Parent. |
| 34 | 4. **Normal Streaming**: Regular streaming of new metrics begins. |
| 35 | 5. **Machine Learning Initialization**: ML models are loaded and prepared for Anomaly Detection. |
| 36 | 6. **Health Check Initialization**: The health engine starts monitoring metrics and triggering alerts. |
| 37 | |
| 38 | Additional considerations: |
| 39 | |
| 40 | - **Compression Optimization**: The compression algorithm learns data patterns to optimize compression ratios. |
| 41 | - **Database Optimization**: The Database engine adjusts page sizes for efficient disk I/O. |
| 42 | |
| 43 | These initial tasks can temporarily increase resource usage, but the impact typically diminishes as the Parent stabilizes and enters a steady-state operation. |