Grafana assumes the scrape interval as 15s by default but COS stack assumes 1m. Without telling Grafana that COS Prometheus uses 1m as the global scrape interval, many graphs will break since $__rate_interval won't be calculated properly.
Context
Hardcoding 1m itself is in question, but this patch fixes many graphs using $__rate_interval without touching the fundamental pieces.
Deploy the COS stack and grafana-agent machine charm and relate those. Open the node-exporter dashboard in Grafana and zoom into 15 min and confirm the graphs are shown instead of "no data".
Also, double check the scrape interval is set as 1m in the data source configuration.
Issue
Closes: #543
Solution
Grafana assumes the scrape interval as 15s by default but COS stack assumes 1m. Without telling Grafana that COS Prometheus uses 1m as the global scrape interval, many graphs will break since $__rate_interval won't be calculated properly.
Context
Hardcoding 1m itself is in question, but this patch fixes many graphs using $__rate_interval without touching the fundamental pieces.
ref: https://grafana.com/docs/grafana/latest/datasources/prometheus/configure-prometheus-data-source/#interval-behavior https://grafana.com/blog/2020/09/28/new-in-grafana-7.2-__rate_interval-for-prometheus-rate-queries-that-just-work/ https://github.com/canonical/prometheus-k8s-operator/issues/544
Testing Instructions
Deploy the COS stack and grafana-agent machine charm and relate those. Open the node-exporter dashboard in Grafana and zoom into 15 min and confirm the graphs are shown instead of "no data".
Also, double check the scrape interval is set as 1m in the data source configuration.
Release Notes