Monitoring Ceph with Prometheus
summary
Shows how to run Prometheus against Ceph using the Ceph manager exporter, node exporter and file_sd for dynamic targets, and how to correlate labels to join OSD metadata with SMART device metrics. Demonstrates PromQL uses like per-second rates and predict_linear to warn if an OSD will fill within an hour, covers alertmanager clustering for deduplication, claims Prometheus scales to tens of thousands of nodes and estimates ~100 MB per node per week of storage.
AI-written summary · about
chapters (8)
- 0:00Introduction to Prometheus
- 2:54Prometheus architecture overview
- 8:14Service discovery in Prometheus
- 11:12Scaling Prometheus monitoring
- 13:38PromQL query language basics
- 20:04Advanced PromQL queries
- 24:52Alerting with Prometheus
- 27:48Q&A and hardware requirements
related talks
31mMonitoring and Centralized Logging in Ceph
38mEnhancing Ceph Monitoring at CERN: Our Approach and Solutions
37mMonitoring Kubernetes with Prometheus
02h 43mPrometheus from A to Z
19mKeynote: Monitoring Kubernetes Clusters with Prometheus
10mPrometheus for Monitoring Metrics - Database Monitoring Tools
32mDemocratizing Application Monitoring with Prometheus
08mPrometheus in 5 minutes