Kubernetes App Metrics

After selecting your namespace and container you get a wealth of metrics like request rate, error rate, response times, pod count, cpu and memory usage. You can view cpu and memory usage in a variety of ways, compared to the limit, compared to the request, per pod, average per pod, etc.

Overview
Revisions
Reviews
Edit →

Run Prometheus in your Kubernetes cluster with a configuration like shown at https://github.com/prometheus/prometheus/blob/master/documentation/examples/prometheus-kubernetes.yml. This will make the data used by this dashboard available.

The haproxy stats are exported by running a sidecar container in the haproxy pod. The nginx stats originate from the nginx lua exporter, which runs in the nginx container itself.