metrics-stack

Configure Prometheus service discovery for Kubernetes pods by defining scrape jobs in prometheus.yml and applying relabeling rules to filter targets. This Skill provides detailed configuration examples for Kubernetes, file-based, and EC2 service discovery setups.

The best way to write PromQL queries for histograms and counters involves using specific aggregation functions to calculate rates and percentiles. This Skill offers guidance on essential PromQL queries for complex scenarios, gauges, and aggregations.

Design Grafana dashboards using the RED and USE methods by selecting appropriate panel types and applying templating to visualize resource metrics. This Skill covers best practices for dashboard design and SLO implementation.

You need long-term storage solutions like Thanos or Cortex for Prometheus when retaining metrics beyond local disk capacity or building highly available clusters. This Skill addresses configuring these solutions for robust monitoring at scale.

Set up Alertmanager and PrometheusRule CRDs for alerting by defining routing rules and threshold conditions within your cluster. This Skill provides configuration examples for Alertmanager and PrometheusRule CRDs to handle critical conditions.

Manage high cardinality and avoid anti-patterns in Prometheus by limiting label values and optimizing query structures to prevent memory overload. This Skill addresses cardinality management and common anti-patterns for robust monitoring.