Grafana MCP for Infrastructure

This will show case how we can host the Grafana MCP, manage authentication when deployed alongside a kube-prometheus-stack operator for incident investigation/postmortoms and remidation.

Kube promethues stack creates promethues which is a timeseries database we can use to record resource metrics, error rates etc, when combined with loki we can have an agent parse logs and even bring up the URL for the dashboard. kube-prometheus-stack also creates promethes rules which creates alerts for things like failed pods/jobs/out of storage etc.