Cutting Through Metrics Cardinality Noise with VictoriaMetrics
summary
Explains why high metric cardinality drives CPU, memory, storage and query latency, shows a live cardinality explorer demo in the VictoriaMetrics UI that lists the top metric names by unique series and the offending label, and points to a Grafana dashboard tracking spikes using the VM new time series created_total metric and a churn rate query. Describes the pipeline components (VM Agent, VMInsert, VMStorage, VMSelect), demonstrates label dropping and VM Agent filtering, and recommends stream aggregation or pre-aggregation when labels cannot be removed.
AI-written summary · about
chapters (10)
- 0:00Introduction and Speaker Background
- 1:06Agenda Overview
- 1:38Understanding High Cardinality
- 3:05Prometheus and Metrics Collection
- 4:45Strategies to Manage High Cardinality
- 8:13Victoria Metrics Overview
- 10:46Cardinality Explorer Demo
- 13:21Scaling and Performance Considerations
- 15:54Conclusion and Q&A
- 18:03Closing Remarks and Lunch Break
related talks
28mThe Sixth Sense: Finding “Dead” Metrics and Silent Alerts in Your Monitoring
28mHow to reduce expenses on monitoring with VictoriaMetrics
40mScale Your Monitoring Solution With the VictoriaMetrics Ecosystem
05mLightning Talk: How to detect and delete unused metrics?
36mVictoriaMetrics: scaling to 100 million metrics per second (EN)
24mManage metrics & logging costs with Grafana Cloud + Log Volume Explorer demo
35mObservability’s Sixth Sense: Detecting Anomalies in Metrics
40mObservability’s Sixth Sense: Detecting Anomalies in Metrics