feat: add K8s SRE Incident Response dashboard

[infrawithshobhit]

Dashboard: K8s SRE Incident Response

A Kubernetes dashboard built for on-call triage, Based on real life example

Q- Why this dashboard?
Most K8s dashboards show everything, during an active incident. That's called noise.
This one surfaces 8 signals that answer, what is broken right now, and why?

Panels included

• Pod Restart Rate — early crash loop detection
• Node CPU Pressure — scheduling risk threshold (warn: 80%, crit: 90%)
• Node Memory Pressure — OOM risk (warn: 85%, crit: 95%)
• Pending Pods — scheduling failure indicator
• Container Restarts / OOMKills — memory leak signal
• p99 Latency — primary SLI for SLO tracking
• 5XX Error Rate — error budget burn rate
• PVC Storage Usage — proactive storage incident prevention

Data source

OpenTelemetry Collector with Kubernetes receiver + kubelet metrics.
APM panels require OTel SDK instrumentation.

Variables

• k8s_cluster_name — cluster selector
• k8s_namespace_name — multi-select namespace filter